After reading Chapters 5 and 6 of Clearing the Air, please discuss your thoughts on finding the balance between reduced CO2 emissions and pollutants such as NOx and PMs. What technologies do you think we should be improving, not just for individual vehicles, but also for major transportation/shipping industries. What options do you believe are sustainable? In grading, I look for thoughtful and unique answers. If you state an opinion or fact, please back it up. You are also graded on driving the discussion and participating with others. Rubric for grading discussions:

• Interacts with other participant’s posts – 2.5 points
• Complete sentences – 2.5 points
• Answers the questions – 2.5 points
• Comments drive discussion/are comprehensive – 2.5 points
• Total – 10 points 

CHAPTER FIVE

The Dash for Diesel

At the Millbrook emissions testing ground, 40 miles north of London, security guards take my phone and laptop and cover the cameras in thick red tape. As a journalist I’ve visited many security-sensitive sites before, from the Houses of Parliament to nuclear power stations, but this is a first for me. When I clear security and step through the gates, I find out why. Paranoia here is high, and so is the amount of money on display. The Millbrook Proving Ground, to give it its full name, is a rite of passage for any new car to make it from a mere prototype to a real-world, production-line model. It is the size of a small town, with emissions testing labs, crash-test centres, noise chambers, atmospheric chambers and 70km (43 miles) of test tracks. Logo-less cars drive around painted in zebra-like stripes to camouflage them from prying lenses, were I to surreptitiously remove a camera sticker.

Millbrook is an independent business – all its activities are funded by the car companies (or, as they call them in the industry, ‘OEMs’ – Original Equipment Manufacturers) who test their new cars and engines here and pay handsomely for the privilege. By the side of one test track I spot a gleaming Aston Martin showroom nestled between the trees. Its high-net-worth customers can test drive their bespoke vehicles on the same track where James Bond sequences have been filmed.

I was invited to visit Millbrook alongside members of the London Assembly – the elected body that scrutinises and advises the Mayor’s Office – here to learn more about the emissions testing process and check up on the fleet of London’s famous red buses that are tested and accredited here. And I, as the only member of the press tagging along, am given a surprisingly long leash.

When Phil Stones, Millbrook’s head of emissions and fuel economy, sits us down for a presentation, two things raise eyebrows. First, nitrogen dioxide (NO2), the EU legal limit of which is regularly breached in London, is not part of the Euro vehicles emissions test. NOx (all oxides of nitrogen combined) is, but the proportion of nitrogen dioxide within that can vary – and the proportion, Phil tells me later, is going up. Second, PMs from tyre and brake wear, despite being known to form a significant percentage of overall PM from traffic, are also not covered by the Euro standards, and therefore not of interest to the manufacturers. ‘There is no legal standard – no legal requirement,’ says Phil. He adds, not for the last time that day, that manufacturers will only build to meet the regulatory standards, and no more.

When it comes to PM2.5 and nitrogen dioxide within towns and cities, by far the greatest single source is the modern automobile. Airparif’s annual reports on the pollution in Paris include ‘hotspot’ maps for nitrogen dioxide and PM2.5, showing red for high pollution, down through yellow and then green for low pollution. On the maps, the city’s roads glow bright red against a yellow-green background. These markings clearly show that urban emissions originate from the roads. In maps that zoom out to show the wider Île-de-France region, Paris is a blotch of red in a sea of dark green, with the occasional yellow vein of a motorway running across it. Traffic accounts for 65 per cent of the NOx and over half the PM2.5 inside the city of Paris. And the vast majority of that comes from diesel. In Paris, diesel vehicles account for about 50 per cent of the traffic, but 94 per cent of the NOx and 96 per cent of the transport-derived PM10.

The exhaust from internal combustion engines produces two kinds of particles: secondary and primary. But the ratio and total number differ, depending on whether it is

a diesel or petrol engine. Professor Simone Hochgreb, formerly of Princeton and MIT and now at the University of Cambridge works almost exclusively on engine emissions: ‘Suppose you start with 100 per cent of fuel, and [say] only 98 per cent of it burns,’ she explains. ‘This means you end up with 2 per cent coming out either as the original fuel or as a messy mix of partly burned stuff, VOCs or particulate matter … This is what’s called “incomplete combustion”. In gasoline engines, incomplete combustion arises primarily from fuel near walls and crevices near the piston when the engine is warm, and from spraying a bit too much fuel when it is cold so that the engine starts. In diesel engines, it arises because you inject the fuel separately from air, and depending on the operating conditions, there is not enough mixing or not enough time to completely convert it into CO2.’ The diesel engine and the gasoline engine therefore work on two very different principles. ‘There is a trade-off in NOx and PM for most types of engines, certainly for diesels,’ says Hochgreb. ‘If one goes down, the other goes up … the simple idea is that a diesel engine is very efficient because it can operate at very high pressures, which also produce high temperatures, which also produce NOx. But in order to use those high pressures and temperatures, you can’t inject the fuel too early because otherwise it will go boom. So, in diesel engines you inject when the air is hot – but now the fuel is not completely mixed with the air, so it will not burn completely, and it will produce the PM, the soot, which comes from incomplete combustion … the PM level is

bad news.’

* * *

I was at a rural wedding in Dorset, southern England, when the Volkswagen diesel emissions scandal broke in September 2015. During the evening reception, in a marquee set in an apple orchard, I slipped outside for a breath of fresh air. I got into conversation with an uncle of the groom, which all too quickly strayed into work. ‘You write for the Financial Times?’ he said. ‘What did you make of this morning’s front page – Volkswagen, eh? It’s got to be the biggest scandal since Enron.’ I hadn’t read that or any paper that morning – I had been trying, with varying degrees of success, to entertain my now 18-month-old daughter on the five-hour train journey to the wedding. So, he filled me in. VW had been caught falsifying the emissions data from its diesel cars, cheating test results to sneak under the radar of increasingly stringent regulations and appeal to eco-conscious consumers. As he spoke I gazed at the bonfire that flickered in the orchard, and thought of my mum’s diesel VW Polo, and the low-carbon credentials she’d mentioned when she bought it.

But the story of how diesel came to dominate our roads began long before a VW engineer said: ‘Hey, why don’t we just cheat?!’ A couple of decades before that malevolent lightbulb moment, ill-considered government policy had already released untold tonnes of diesel fumes into the air.

The Kyoto Protocol in 1992 required governments to reduce CO2 emissions by 8 per cent from 1997 to 2013. Given the known global warming effects of CO2, this was a very necessary step. The means of reduction, however, wasn’t stipulated, and in Europe it led to the widespread adoption of diesel vehicles. Due to its relative fuel efficiency diesel can produce 15 per cent less CO2 than petrol engines. The car industry, sniffing an opportunity to build vast amounts of new cars, lobbied the European Commission to promote diesel. It was pushing on an open door. In 1998 the EC issued a commitment to cut CO2 emissions by 25 per cent in all new cars sold within ten years; the only available means of doing so was a switch to diesel engines. Most EU countries began to introduce tax breaks to incentivise consumers to buy diesel cars over petrol. In 2001–2 the UK began taxing vehicles according to CO2 emissions. Cars with lower CO2 emissions fell into cheaper vehicle excise duty (car tax) bands, which gave diesels a cost advantage, and fired the starting gun for what became known as the ‘dash for diesel’.

It was a case study in effective government policy. With similar tax and fuel incentives rolled out Europe-wide to meet the EC ruling, the market share for diesel cars across the continent rose from under 10 per cent in 1995 to over 50 per cent by 2012, while the share of diesel in total fuel consumed reached 63 per cent. From 2001 to 2010, the proportion of diesel among all new cars registered in Norway rose from 13.3 per cent to 73.9 per cent and in Ireland from 12 per cent to 62.3 per cent. Buoyed by such success, European carmakers took their diesel models global. In India, diesel went from just 4 per cent of new car sales in 2000 to half of new car sales by 2016. When I visited the Central Road Research Institute (CRRI) in Delhi, Dr Niraj Sharma told me that diesel was initially highly subsidised for agricultural use:*

‘It used to be that 90 to 92 per cent of all vehicles used to be petrol driven, and around 8 to 9 per cent were diesel,’ he says. ‘Now things have changed … diesel vehicles started dominating … the subsidy that was meant to be for the farmers, is mostly being used by the automobile manufacturers.’ Diesel on the streets of Delhi, he says, emits ‘more hazardous air pollution than the petrol-driven vehicles’.

In 2000, when these incentives started, even the most efficient diesel cars emitted over three times more NO2 per kilometre, and ten times more PM. Most cars on the road were much older, emitting at least four times more NOx and between 22 and 100 times more PM than petrol engines. A conscious trade-off was therefore made by policy-makers to accept ill-health as a result of increasing pollution in our towns and cities, in order to meet CO2 targets. Simon Birkett, a former London City banker who gave up his job to set up the campaign group Clean Air London in 2007, told a Guardian exposé in 2015 that ‘even though the European commission, national governments and the car industry knew how dangerous diesel is, together they incentivised it and deliberately engineered a massive switch away from petrol – without any public debate.’ The article further quoted a retired ‘very senior’ civil servant who recalled ‘the health issue’ as being a significant factor in departmental debate at the time: ‘We did not sleepwalk into this … everyone had to swallow hard.’1

Any claims that ‘we didn’t know how bad diesel was’ ring extremely hollow. The International Agency for Research on Cancer (IARC) had named diesel exhaust as a probable carcinogen back in the 1980s. In 1986, Dr Robin Russell-Jones, a lung expert who successfully campaigned against leaded petrol, gave evidence to a UK House of Lords select committee that diesel pollution was linked with asthma, cardiovascular disease and lung cancer. In 1993 a major report for the Department of the Environment by the Quality of Urban Air Review Group (QUARG) stated that diesel emissions were ‘a potential health hazard’, containing ‘compounds known to be carcinogenic [that] may cause impairment of respiratory functions … an increase in mortality and morbidity may be associated.’ In 1996, a POST (Parliamentary Office of Science and Technology) report given to UK Members of Parliament to brief them on scientific developments warned of ‘emerging evidence that fine particles in the air could be a significant contributor to respiratory disease and death … fine particles from diesels and other sources may contribute to significant mortality around the world.’ It continues, ‘road transport is the biggest single source of particulates [and] diesel emissions are the dominant source … There is also evidence linking exposure to diesel exhaust with higher rates of lung cancer.’ The POST report again drew attention to the earlier findings of the QUARG report, just in case anyone missed it the first time, that any increase in the market share of diesel – which was around 20 per cent at the time – would ‘inevitably make matters worse as the current technology diesel car emits far more particulate matter than the modern petrol car’ with ‘no current prospect of the diesel improving beyond the petrol car’.2 When I read that, I genuinely got goose bumps because the information was there, in black and white. Ignorance was not a defence. But no one heeded the warning.

Peter Brimblecombe, who sat on the QUARG committee at the time, clearly remembers that ‘the message of the group was that particles are the big problem and they’re diesel particles.’ Records received by the BBC in November 2017, after a two-year freedom of information battle, confirmed that ministers and civil servants in the government were well aware that diesel pollution was bad for air quality. Advice from the Treasury’s tax policy section presented to ministers – the most senior members of the government – unambiguously stated: ‘Relative to petrol, diesel has lower emissions of CO2 but higher emissions of the particulates and pollutants which damage local air quality.’

Governments across Europe ignored all this. Promotions for diesel as the ‘green’ option continued unabated. There was some (very slight) method to this madness, however: the Euro emissions standards, which all new cars had to pass before sale in the EU, promised that cleaner diesel was just around the corner. The Euro 1 emissions standard began in 1992, requiring all new cars sold within the EC from that point to meet certain emissions criteria for NOx, PM, CO and hydrocarbons. The bar started off high, but the plan was always to bring it down a notch or two with each new emissions standard: Euro 2 came in 1996, followed by Euro 3 in 2000 and Euro 4 in 2005 (we are, at the time of writing, currently at Euro 6, introduced in September 2014). Due to the inherent differences in engine technology, diesel was allowed to emit higher amounts of NOx and PM than petrol, but the gap was due to get ever smaller. The car companies could repeatedly say, ‘yes the emissions are high now, but just wait till you see our next model’.

Frank Kelly, who has been part of the government advisory Committee on the Medical Effects of Air Pollutants since the early 2000s, admits: ‘There were reports saying, “if you make this decision based on benefit from CO2 decreases there is the likelihood that there will be increases in PM and NO2 concentrations in urban areas’’. But I think in their defence – and there’s not a lot of defence, because there wasn’t a lot of holistic thinking going on at the time – it really was [about tackling] climate change and CO2. At the time the Euro standard – they were at Euro 3 then I think – was the big plan [to] increase emission control.’ Yet the Euro 3 standard for diesel cars had a 500mg/km limit value for NOx compared to 150mg/km for petrol, and 50mg/km PM compared to just 5mg/km for petrol. If there was a conviction that the Euro standards would eventually save us, there was an equal acceptance that it was OK to expose the populace to increasingly poisoned air in the meantime.

In little over a decade, the number of diesel cars – all legally emitting several times more NOx and PM than the petrol cars they replaced – rose from under 2 million in the UK to over 12 million. James Thornton, the CEO of ClientEarth, a non-profit environmental law organisation, remembers appearing on a TV panel discussion when the VW story broke, alongside Sir David King, the government’s former Chief Scientific Advisor from 2000 to 2007: ‘He was saying, yes, we did believe [diesel] was lower CO2 emissions … but we were hoodwinked by the car companies, we didn’t realise the emissions were nearly as bad as they were or would be because the car companies promised us it was very easy to stick on filters, collect the particles and otherwise reduce dangerous emissions – and that didn’t happen.’ If the Euro standards were adhered to, ‘they might have been a decent compromise’, says Thornton. ‘But they haven’t been, so they were a terrible compromise.’

Diesel vehicles have consistently been found to fail the Euro emissions standards when tested in the real world. Researchers at King’s College London undertook their own testing of over 80,000 vehicles at roadside locations in 2011 and found there had been little or no improvement in NOx emissions from diesel cars, vans, HGVs or buses for over 20 years (it did find a significant improvement for petrol cars), despite the Euro standards in place to reduce NOx. The study estimated that even the Euro 5 diesel cars, the best available at the time, emitted over 1.1g of NOx per km when on the road – more than five times the Euro 5 limit of 0.18g/km, and more than even the original Euro 1 ‘high bar’ limit of 0.97g/km set in 1992. Other studies began appearing with similar findings. The European Joint Research Centre found that petrol cars largely performed within Euro emissions limits, but diesel cars emitted levels 4 to 7 times higher than they were supposed to. In 2010, when around 3.6 million Île-de-France (Greater Paris region) residents were exposed to NO2 levels exceeding the

annual limit, Airparif reported that, ‘Although the filters that now equip most new diesel vehicles contribute to reducing particulate emissions, they also give rise to a significant increase in NO2 emissions. It is now confirmed that the proportion of NO2 in NOx emissions is increasing steadily.’ In 2012, half of all private cars in London

and virtually every bus, HGV, LGV and black cab, were running on diesel. That same year, just before London hosted the Olympic Games, the IARC upgraded diesel engine exhaust from ‘probably’ carcinogenic to humans to ‘definitely’.

To make matters worse, the CO2 advantage associated with diesels never materialised either. Research suggests that the global warming impact of increased black carbon emissions from diesel more than offsets the CO2 saving, due to black carbon’s ability to absorb and radiate heat. NOx also includes nitrous oxide (N2O), which is a more potent greenhouse gas than CO2, and diesel emits higher levels of NOx than petrol. Not just that, but many Euro 6 petrol cars now achieve almost the same fuel efficiency – and therefore CO2 emissions per kilometre – as diesel. Europe gained all the ill-health for none of the climate benefits. And there were alternative options. To meet the same Kyoto targets, countries such as Japan, South Korea and the US chose to back research into low-emission hybrid and electric vehicles. Diesel in the US has always been ‘socially and environmentally unacceptable’, says Hochgreb. ‘Why? Because European diesel fuel is a lot cleaner, much more controlled in terms of quality. The US fuel was not. It didn’t work in the US because there were [comparatively] no regulations on fuel … of course all of the freight uses diesel, terrible diesel, and terrible engines. But in Europe the incentives were there, there was a demand for higher efficiency [than offered by petrol] because of fuel prices and high taxes on fuel, much higher than the US … [The market share of diesel in Europe] went to 50 per cent of cars because it was perceived as environmentally friendly.’

And only after all that, comes the VW scandal. The success of Volkswagen’s diesel cars was based almost entirely on its environmental credentials. Riding high after the decade-long dash for diesel in Europe, VW tried to break the US market. It ran a commercial during the 2010 Super Bowl of its diesel Audi stopping by a long line of grey, wheezing, smoky cars, before being waved through by the ‘green police’ saying ‘You’re good to go.’

An in-depth description of the VW scandal in Fortune, March 2016, outlined Volkswagen’s ‘audacious’ goal to become the biggest car company in the world. They saw cracking the US market as ‘crucial to the mission’. However, California’s emissions rules were getting in the way.†

Chancellor Angela Merkel personally weighed in on the issue in April 2010 when she took on the head of the California Air Resources Board (CARB), Mary Nichols, in a private meeting. However, Merkel picked a fight with the wrong person. And given that the third person in the room was Arnold Schwarzenegger, that’s really saying something. Mary Nichols, chair of CARB from 1979 to 1983, had returned for a further spell in charge – at the request of Governor Schwarzenegger – in 2003, and has stayed in that position ever since. She’s rightly been described as a ‘rock star’ of the air quality world, and a known tough negotiator. When I spoke to her over the phone from New York, I asked if it was true that Angela Merkel asked her in that meeting to relax the NOx limits for the sake of the German auto industry? ‘I actually had to testify on this under oath in the German Bundestag,’ she laughs. ‘She didn’t ask me to do anything, what she said was – in front of my then boss, Governor Schwarzenegger – “Your diesel emissions standards are too severe and they are hurting our German companies”. It was an accusation, not a question, I guess [in effect] “you are doing something wrong, you should stop!”’ And what did you

say? I asked. ‘I said “I don’t think it’s true. We need to have these controls because of the need to meet health standards.” I answered back, being that sort of person! … We knew already that the German companies were opposed to our NOx standards because we met with them all the time. But we weren’t expecting [to hear] these things from the Chancellor.’

In Europe too, the Euro regulations were getting tighter. The PM emissions allowed for diesel passenger cars had gone down from 25mg/km in the Euro 3 regime to just 5mg/km in Euro 4 and 4.5mg/km in Euro 5, the same as petrol. The gap between diesel and petrol cars for NOx emissions was also closing. Whereas Euro 5 allowed diesel cars to emit up to 180mg/km of NOx and petrol cars 60mg/km, from September 2014 Euro 6 would allow just 80mg/km for diesel, while petrol would stay at 60mg/km. A 2014 EU report anticipating the changes concluded, painfully ironically in hindsight: ‘The worry is that real-world emissions might not show the same decrease.’

Regulators have historically relied on highly controlled lab-based emissions tests at facilities like Millbrook. An automotive company brings their latest model in for testing, and the ‘drive’ takes place on rollers in the lab (known as ‘dynos’), at an exact temperature and speed, ensuring that the test is precisely the same and repeatable for every model. You compare like for like, cancelling out all other variables. At Millbrook I got to watch one of these tests in action from the control room, viewing the car through a small, thick-glass window. A shiny new SUV is perched seemingly stationary while its wheels whirl around on the dyno below; a professional driver inside it slowly decelerates down from 80kmph to zero. In the next room a bag, like a giant sandwich zip bag, slowly fills with the exhaust fumes. I ask if I can touch it, and I feel the warmth of the smoke as it slowly inflates. The exhaust from the bag is then passed through various analysers to measure for levels of NOx, carbon monoxide and PM.

Unfortunately, as Fortune put it, ‘that approach makes it possible to cheat’. Software now known as a ‘defeat device’ could be installed to recognise when the car was within lab conditions, and ensure the car only emitted what it was allowed to during the test and not a minute longer. The one installed by VW was found to increase NOx emissions to a factor of 10 to 40 times above EPA (Environmental Protection Agency) compliant levels as soon as it was out of lab test conditions.

In 2013, the US-based non-profit International Council on Clean Transportation (ICCT) began looking into the disparity between the claimed performance of European diesel cars and their real-world emissions, finding NOx emissions of up to 35 times permitted levels. In May 2014 it sent its report to the EPA. CARB and the EPA, both suspecting a defeat device, began scrutinising VW’s diesel cars in minute detail for most of the following year. ‘The ICCT organisation actually built the case, and brought it to us,’ confirms Mary Nichols. ‘The EPA joined with us in the prosecution of it.’ Aware they were about to be found out, VW privately told the EPA on 3 September 2015 that their cars were indeed fitted with illegal software, perhaps hoping for a behind-the-scenes slap on the wrist. The EPA instead revealed VW’s cheating via a public ‘Notice of Violation’ issued on 18 September 2015, stating unambiguously that ‘VW manufactured and installed defeat devices in certain model year 2009 through 2015 diesel light-duty vehicles … These defeat devices bypass, defeat, or render inoperative elements of the vehicles’ emission control system …’ In the six-page public notice, the EPA also reiterated that the purpose of having emissions standards was ‘to protect human health and the environment’. Within days, VW admitted it had installed defeat devices in up to 11 million cars worldwide, beginning in 2008.

‘I was shocked by how long it had gone on,’ admits Mary. ‘Frankly, as an appointed official in the government I was concerned that we would be seen as, and perhaps we really were, at fault for not having found it sooner. I was angry about the cheating and wanted to make sure that we got it to stop, and punished the violations, but I was also worried that … it might be a much more widespread problem.’

I asked Phil Stones at Millbrook how VW had managed to pull the wool over his eyes, and that of other testing facilities, for so long: ‘The regulations are [that] what it does on the dyno it must do on the road, in the same conditions, or words to that effect. So, if you drive on the dyno, same conditions for me – this is my interpretation – is that it’s that speed load, it’s that time, it’s that same window of operation.’ The VW software could tell when the car was in test conditions because the car was running and the wheels were spinning, but the steering wheel wasn’t moving: ‘It doesn’t see any steering input on the dyno, so it went to a mode that “defeated” [the emissions].’ VW cars were able to meet the Euro limits, but only for long enough to pass a test when no steering is registered on the steering wheel. On the road, as soon as the steering wheel moved, the VW could effectively loosen its belt and exhale high levels of pollutants once more. I asked Phil what the atmosphere was like in Millbrook the day after the scandal broke in September 2015. Millbrook staff work alongside engineers from all the major car companies every day – if they are not there for regulatory tests, then they hire out Millbrook facilities to test out prototypes. Although he’s not allowed to confirm it, Phil most likely worked alongside VW engineers who knew their models were cheating, and watched as Phil and his team ran the tests. ‘Everybody was surprised that they knowingly, directly used a defeat device,’ says Phil. ‘People “optimising” is different, it happens in any industry, sport, Formula One, anything – people optimise and make decisions commercially. But cheating puts them in a different world. It’s like a sportsman – does he do high-altitude training? Or does he take steroids? That’s the difference in my world. They took the steroid drug.’

Andy Eastlake, now managing director of the Low Carbon Vehicle Partnership (LowCVP), was a senior emissions engineer – and Phil Stone’s former boss – at Millbrook, until 2011. ‘For many years we were focused on air quality through the Euro standards, and all of the [air quality] modelling that we put in place assumed that we would deliver to those standards,’ he tells me. ‘I remember presenting on “real-world testing” papers back in 1997, 2000. Real-world testing is not a new issue. What has changed, and it was genuinely a shock to me, was the blatant fraud, if you like, against those regulations. Everyone knew that those regulations weren’t necessarily as all-encompassing as they should be … the risks of only testing a small area of operation and assuming it will be clean for the rest of the operation – the risk had gone up manifestly. And then you’ve got an individual company that has driven a coach and horses through that loophole.’

VW’s was very, very far from being a victimless crime. National and city transport decisions are based on forecasted emissions data; the falsifying of that data led to many thousands of deaths. According to one analysis, the 11 million VW cars fitted with defeat devices would have collectively emitted close to 1 million tonnes more of nitrogen dioxide every year than the policy-makers, regulators or car owners were expecting. And according to the EEA (European Environment Agency), nitrogen dioxide pollution causes 78,000 premature deaths in Europe every year. Paul Bate was a senior transport engineer for Derby City Council from 2001 to 2007. ‘I can remember the nice graph that said “this is [the emissions from] Euro 4, and this Euro 5, and by the time get to Euro 6 there is no [air quality] problem whatsoever”. And what happened is we got to a certain

point and then emissions flatlined … vehicles were being optimised to pass a particular test, whereas in the real world they were performing sub-optimally. So what we ended up with is a disconnect in the information that planners in city authorities across the world have been given to plan for and improve air quality in their areas, [it] has been wrong by several orders of magnitude … the standards, and all the tools and models, are based on the fact that vehicles perform according to the Euro standards. And there was a very real difference.’

In the weeks after the VW scandal, Ally Lewis and Frank Kelly penned a joint letter to Nature, stating that ‘pollution from diesel vehicles has long been under-reported’. When I met Kelly I asked him whether they had suspected foul play. ‘From the year 2000 we had predictions for what NO2 would do in London, for example. And the predictions were all very favourable. There was a line going down. Our measurements, however, from 2000 to 2006–7, were a straight [horizontal] line. What we were predicting was going to happen to air quality was getting increasingly distant from what was actually happening. So, we started scratching our heads around 2005 to try to understand … We ended up getting in some laser-based equipment from America: you set it up at the side of the road, shine the laser across, and whenever a vehicle goes past it tells you what the NOx is in the vehicle exhaust plume.’ After tens of thousands of different vehicles had passed the laser, Kelly’s team found that what the vehicle was meant to be doing was nothing like what the vehicle was doing: ‘We got the equipment from the university of Denver with funding from Defra [the UK Department for Environment, Food and Rural Affairs], and they got our report around 2007. And [Defra] basically sat on it. So, we did some more work organised by the manufacturer, showing that certain cars and vehicles were a lot worse and some a lot better than others, and [Defra] got that information as well. Clearly, they were uncomfortable about reporting this because it would bring certain

car manufacturers into a bad light. We knew what the problem was, we had the data, and Defra had the data, long before – a couple of years before – the [VW scandal] broke in the States … it took the Americans to actually do something about it.’

There was, agrees James Thornton of ClientEarth, a ‘grand collusion of most of the motor industry to delude the regulators and the public, although the regulators did know about quite a bit of it and did nothing’. He repeats Kelly’s allegation that, ‘in the UK the relevant agency did know and didn’t do anything, it took the US Environmental Protection Agency to blow the whistle … the real problem has been the huge investment of a German car company in diesel engines and the control that the German car companies have over the German economy and therefore over a lot of the European economy. The reason Volkswagen diesel was eventually discovered was that Volkswagen had invested … hundreds of millions in developing their new generation diesel engine and having spent so much on it they decided it was time for America to fall in love with diesel. So, they took it to the US and they neglected to understand they didn’t have the same sweetheart deal with the US EPA, who actually tested it and discovered that they had these defeat devices.’‡

By the time VW were at it, however, so were many of the other car companies. Phil Stones was happy to tell me about a few more: ‘The Fiat one was a timer of 22 minutes. The test is for 20 minutes. Their defence was [subsequently] “well on the road, it does exactly what it does on the dyno”. The dyno only goes for 20 minutes, so they defined that that’s how long the regulators want it compliant for [and] after that they can do what they like. So they did. They “modulated”, as the Italian Transport Minister said, they modulated their emissions – their after-treatment – down, after 22 minutes … Others put temperature boundaries in. The test was between 20 and 30 [degrees Celsius] – so on the road, between 20 and 30 degrees, it did exactly what it did on the dyno. Drop to 17°C, it didn’t do the same.’ Because everyone was attempting to hoodwink the regulators, does it suggest that the Euro 5 or 6 regulations for diesel were becoming unachievable? ‘Er, no. They were achievable,’ says Phil. ‘My view is that in a commercial world you will do as much as you need to do relative to your competitor.’§

An independent ‘real-world emissions’ league table, called the EQUA Air Quality Index, overseen by a number of academics from institutions including King’s College and Cambridge, emerged following the VW scandal. It takes all new car models out on the road to see what really comes out of their exhaust pipes. At the time of writing there are eight models of diesel car manufactured to meet the latest Euro 6 standards that not only don’t meet it, but don’t meet any of the previous standards going back to 1993. Two Nissans, a Fiat, Subaru, Peugeot, Renault, Infiniti and Ssangyong (no, me neither), were all rated ‘H’, meaning ‘No comparable Euro standard: roughly equal to [polluting over] 12+ times Euro 6 limit’. If you include models rated G ‘No comparable Euro standard: roughly equal to 8–12 times Euro 6 limit’ and F ‘No comparable Euro standard: roughly equal to 6–8 times Euro 6 limit’, of cars launched after the Euro 6 requirements of 2014, there are another 29 models, by manufacturers including BMW, Ford, Mercedes-Benz, Volvo and Vauxhall, to name just a few. If you add in all supposedly Euro 6-standard cars that only managed to reach Euro 3 rating (a standard that’s fourteen years out of date), Euro 4 (nine years out of date) or Euro 5 (three years out of date) during EQUA’s road test, then there are a total of 134 diesel car models, from virtually any car brand or vehicle size you could name: all met the official lab-based Euro 6 emissions test and were released onto the market, and yet all fall short when independently tested in real-world conditions. And how many Euro 6 diesels did actually manage to meet the Euro 6 limits during the EQUA road test? Just ten (and, to give them some credit, six of them were made by Volkswagen).

We arguably have VW to thank for raising diesel emissions so high up the political agenda and into public consciousness. Ally Lewis in York even suggests that ‘Dieselgate’ was the year dot for everything that has happened to his field of research since: ‘Knowing the rate of change that was occurring before VW and the rate of change that has occurred after VW … the rate has increased by a factor of 10,’ he says. ‘I suspect VW will go down in history as probably making the single biggest contribution to improvements in air quality in Europe … The 2040 ban on [UK sales of petrol and diesel] cars … that would never have happened without [the VW scandal].’

Volkswagen’s unintended altruism didn’t come cheap, however. In the US it has pleaded guilty to three criminal felony counts and agreed to pay fines totalling $2.8 billion, plus a civil resolution of $1.5 billion. In December 2017, Oliver Schmidt, head of the company’s environmental and engineering office in the US, was sentenced to seven years in prison by US courts. By February 2018, VW was reportedly facing a bill of $25 billion worldwide. In May 2018 Martin Winterkorn, the former CEO of Volkswagen at the time of Dieselgate, was charged by the US authorities with ‘conspiracy to defraud’ and a warrant was issued for his arrest. In June 2018, the chief executive of Volkswagen’s Audi division, Rupert Stadler, was arrested in Germany as part of the emissions scandal investigation. However, because the light of scrutiny soon lit up other car makers, too, VW’s overall market share barely wavered. And the scrutiny spread to encompass air pollution from other transport sectors too. It was the name ‘Dieselgate’ that stuck, not ‘VWgate’, and come 2018 it wasn’t about VW any more, it was about the fuel itself: diesel.

* * *

Trains are undeniably the ‘greener’ option for travel compared to cars. Given that you can fit far more people into them, the average fuel use per passenger will always be lower. But diesel fumes from diesel trains expose passengers and freight workers to a surprising amount of pollution. All commuter trains in 18 out of 26 transit agencies in Canada and the US are hauled by diesel locomotives. About 20 per cent of Europe’s current rail traffic is towed by diesel, with the UK, Greece, Estonia, Latvia and Lithuania around or above 50 per cent. Over half of India’s 20,000-plus trains also run on diesel, consuming around 7.4 million litres of it a day.‖

A month-long emissions study in Seattle found that the average PM2.5 concentration was 6.8mg/m3 higher for people living near railway lines. Inside the train itself, a Boston study in 2010 found the mean PM2.5 reading to be 70mg/m3 in the front carriage and 56mg/m3 in the rear (a separate study in Toronto also found pollution levels to be 3.7 times higher in the front carriage compared to the rear, so note to self: always sit in the rear of a diesel-pulled train).

A 2015 survey of Birmingham New Street station, one of the busiest in the UK, found hourly concentrations of up to 58mg/m3 for PM2·5 and 29mg/m3 for black carbon. The highest levels were associated with idling trains, with concentrations up to six times higher compared to passing trains. Having travelled to Birmingham New Street many times in my life, it has a noticeably low ceiling height, giving the feel of an underground station, with very little space for the air to circulate. Unlike an underground system, where all trains are electric, however, here the majority are diesel trains. Yet we seem to have a blind spot for their emissions, compared to diesel cars. If our everyday commute involved a giant taxi rank in a huge underground tunnel, with all their engines running, the problem would be immediately apparent. New Street station apparently has an impulse fan system that responds to the levels of carbon dioxide in the station: the fans begin to operate once carbon dioxide exceeds 1,000ppm and the speed of the fans increases relative to CO2 levels. However, such fans only respond to carbon dioxide levels, the bare minimum to stop us from suffocating: NO2 and PM levels could in theory continue to rise unabated.

When Cambridge University tested emissions at London’s Paddington station in 2015, around 70 per cent of its trains were powered by diesel engines. It’s a one-way terminus, meaning that trains must enter and exit from the same side, resulting in many ‘idling’ with their engines running rather than switching them off during turnarounds. Around 37 million passengers use it every year, many of whom use it twice every working day. The Cambridge team led by Dr Adam Boies found that Paddington station’s hourly mean PM2.5 hit a maximum 68mg/m3, while NO2 maxed out at 120ppb (which would be a breach of the EU hourly NO2 limit of 105ppb, if the station didn’t have a roof – but it does, so outdoor guidelines don’t apply). Sulphur dioxide (SO2) concentrations, virtually non-existent outside the station, averaged 25ppb. The peak concentrations between 7 a.m. and 10 a.m. corresponded precisely with peak idling train activity.

Despite my repeated interview requests over several months, Network Rail, the operator of train lines and stations in the UK including Paddington, declined to speak to me on the subject of air pollution and diesel. The same happened when I approached a number of private train operators directly, too. Despite being very happy to market themselves as the green way to travel,¶

they get very spiky when it comes to discussing diesel. Regarding the Paddington study, one harassed Network Rail press officer eventually responded by email: ‘I am waiting for a report to be sent over to me by our environmental team which outlines the air pollution levels at Paddington. Once I have this I will share with you but, to give you a flavour of what it contains, I am told that air pollution was not as bad as expected due to good levels of ventilation with the biggest pollution source was Burger King extraction fan.’ When he sent me the report a few days later, its title – ‘Air and health impacts of diesel emissions’ – promise much. Published in January 2016 by the Rail Safety and Standards Board (RSSB), an ‘independent’ UK body with a membership comprised entirely of the UK rail companies, however, it is very obviously, and disappointingly, an industry whitewash. Its opening salvos talk up the role of diesel in UK rail ‘for the foreseeable future’ and play down the health concerns, mentioning just two studies where ‘The carcinogenicity of diesel exhaust emissions appeared to be specific to rats … not relevant to humans’ and ‘no evidence of increased risk at lower cumulative exposure levels’. As we’ve seen in all the previous pages of this book, and will see even more evidence of in Chapter 6, that is total rubbish. Regarding Cambridge’s Paddington study, the RSSB report is a virtuoso performance in deflection tactics: ‘the position of the monitoring equipment suggests that the results may have been influenced by other sources of particulates independent from the railways. One of these was adjacent to the Burger King outlet … attributed to cooking emissions, corresponding to the busiest times for the outlet. The highest particulate concentrations were recorded on the Praed Street ramp, some distance from the platforms, where smokers habitually gather; it would therefore appear that these more local sources of pollution were significantly affecting the overall results.’

I sent the RSSB report to Dr Adam Boies at Cambridge. It was the first time he’d been made aware of it, despite it being out for over a year – clearly the RSSB had never contacted the academics involved to better understand their findings. Understandably, Boies was not impressed: ‘We had set out to measure emissions directly from the trains, but were not allowed to do so by any of the operators,’ he told me. ‘Had we been able to do so, we would have been able to answer the question that is being pointed to [by RSSB]: were the high levels of pollutants in Paddington station a result of the cooking or the train emissions? At the time we conducted some “back-of-the-envelope” estimates of comparisons between likely emission rates between trains and the Burger King at Paddington and found it highly unlikely that the Burger King emissions were greater than the trains … We did conduct chemical analysis of the collected soot samples to determine whether cholesterol (a marker for emissions from cooking) was higher in the particles near the station centre. The results of this analysis showed that there was not a measurable level of cholesterol in the soot samples near the station centre, thus the particles were not identified as being explicitly from cooking. The truth is that we could answer this question quantitatively if the diesel emissions were measured from the fleet of UK diesel trains. We were not allowed to do so.’

Frank Kelly at King’s College London agrees that ‘both NO2 and PM2.5 are coming out of diesel engines in large emissions … it’s recognised that diesel trains need to be replaced … At the moment there’s no discussion around diesel trains affecting air quality and people’s exposure – if we start having that discussion then you never know what might happen … we need to replace diesel trains with electrified trains.’ Instead, in July 2017, the UK Transport Secretary Chris Grayling cancelled the planned electrification of three major UK rail lines.

Another high diesel emitter that lurks under the radar of public consciousness is the diesel generator. Used to produce off-grid electricity, these range from small generator sets used by market stalls to large ones used on construction sites. The smallest class of diesel generator (below 19kW) comprised 18 per cent of the off-road machinery in the United States in 2004, but generated 44 per cent of diesel PM and 12 per cent of NOx emissions derived from mobile sources nationally. According to London’s Atmospheric Emissions Inventory for 2010, non-road mobile machinery including generators accounted for 10 per cent of NOx in greater London and 11 per cent of PM10. An article in the April 2017 edition of

Environmental Scientist magazine headlined ‘Will backup generators be the next “Dieselgate” for the UK?’ asked why no public official log or database of diesel generators existed despite thousands running across the UK, arguing that the ‘air quality impact from the operation of diesel generators has often been overlooked’.3 According to the author, a relatively modest 8 megawatt (MW) ‘genset’ would emit NOx at the rate of 26.7 to 42.2 grams per second.

The use of diesel generators is greater still in developing and middle-income countries where the power supply can be unreliable. With not enough electricity to meet the urban demands of Nepal, one local newspaper reporter describes the familiar sounds that accompany blackouts: ‘thousands of diesel generators rumble to life, spewing noxious particulate matter.’ In Delhi, a local businesswoman who runs a factory with around 200 sewing machinists, told me: ‘Of course there are diesel generators running everywhere … And why? Because our government cannot provide us with 24-hour electricity. You don’t get enough power supply [from the municipal supply], it is only that … they all have 24-hour diesel back-up, everybody has these massive generators. All construction uses diesel generators … You need an uninterrupted power supply. You can’t hire labour and then ask them to sit down and get up, sit down and get up … It all boils down to the basics. Get those basics sorted out, and everything will get better.’

The sector perhaps most addicted to low-grade fuel, however, is shipping. It is easily the transport sector with the dirtiest history. Shipping emissions contribute nearly 15 per cent of NOx and 13 per cent of sulphur dioxide emissions globally, and these numbers are increasing. Due to growing populations and consumer spending, more and more supertankers set sail every year. Since 1985 global container shipping has increased by about 10 per cent annually, with only brief dips for each recession. The Third International Maritime Organization (IMO) Greenhouse Gas study forecasts that by 2050, CO2 emissions from international shipping could grow by between 50 per cent and 250 per cent depending on future economic and energy growth. Crude oil accounts for roughly a quarter of all goods transported by sea. Which gives us the headache-inducing fact that a quarter of all shipping emissions come from shipping the fuel needed to produce the emissions.

In Europe, shipping in the Baltic Sea, the North Sea and the English Channel causes more than 800,000 tonnes of NOx each year (roughly four times the total amount that Belgium emits each year). Shipping fuel also contains up to 3,500 times the sulphur content of the diesel used by road transport. In southern California, says Sam Atwood at the South Coast Air Quality Management District, the twin ports of Los Angeles and Long Beach combined are the largest port complex in North America. ‘Until recently the ships burnt some of the very dirtiest bunker fuel in the world,’ he says, ‘extremely high sulphur content fuel, and then once the ships get to the port the containers are offloaded onto diesel trucks, and then possibly taken to an intermodal centre where they are loaded onto diesel locomotives.’

After my bad experience with the railway authorities, I was pleasantly surprised when Simon Bennett, director of policy at the International Chamber of Shipping, agreed to talk to me. ‘It’s a complex issue,’ he says when we begin. ‘In shipping we’re regulated by the International Maritime Organization, that’s a UN agency, and pretty well all of our regulations are adopted using a global regulatory framework … Shipping is inherently international, you’re moving cargo from one country to another, [so] in simple terms if you had different rules at different ends of the voyage you’d have chaos … for the best part of a hundred years we’ve actually had a framework of global regulations.’ When it comes to environmental regulations, shipping is governed by the MARPOL convention, an international agreement originally established to address disasters such as oil spills, but recently expanded to encompass things like NOx and sulphur dioxide. ‘The mid-1990s [was] the first time it actually set a cap for the sulphur content of marine fuel,’ Bennett informs me, adding ‘it was admittedly quite a high cap.’ I am genuinely astonished by what he tells me next. ‘For the last 30 years or so the majority of ships have actually been burning residual fuel, which is basically the dregs from the oil-refining process, like tarmac … the reason why we’ve been using it is because the oil industry was very keen for the shipping industry to use residual fuel because in simple terms they had nothing else to do with it.’ I ask if ‘residual fuel’ is also known as heavy fuel oil, or HFO? ‘Yeah, HFO … it suited everybody, it’s provided the world economy with cheap goods transport, and as I say the oil industry were very keen to provide this, initially they were almost providing it free of charge … Until about the late 1960s, I guess ships were actually using diesel and then they switched from diesel to residual because it was so much cheaper and ships were also becoming much, much larger as well.’ Since then, millions of tonnes of HFO have been burned in densely populated port cities and the world’s most pristine marine wildernesses. More than 850 ships operating in the Arctic today are thought to use HFO, according to the Clean Arctic Alliance, representing around three-quarters of total Arctic ship fuel use.

The leisure cruise ship industry, which bases its entire image on healthy, clean-living, outdoor experiences, is not much better. According to an annual survey of the biggest cruise ships operating in Europe, conducted by the German environment group NABU (Nature and Biodiversity Conservation Union), a mid-size cruise ship can burn through 150 tonnes of low-grade diesel each day, emitting as much PM as one million cars, as much NOx as 421,000 cars, and the sulphur emissions of 376 million cars (yes, you read that right –

three hundred and seventy six million cars). The report authors accuse the industry of having ‘contempt’ for the health of its customers. I spoke with Sönke Diesener, transport policy officer at NABU, who told me, ‘Shipping in international waters seems to be under the radar, nobody cares. Most ports are far from city centres … Awareness is rising [but] unfortunately shipping is a very slow changing business. The International Maritime Organization is dominated by pro-shipping countries, and small flag-of-convenience-states like Bahamas or Liberia (where ship-owners flag their vessels to save on tax and labour costs) have power to set the agenda.’ According to NABU, just one large cruise ship in port will probably release more air pollutants than all the cars found within most port cities.

As with rail operators, I approached many ports and most didn’t want to speak to me. Most have stuck their head in the (oil-slicked) sand on this issue, believing that ignorance is a preferable option. In 2017, a BBC investigation revealed that Southampton, the biggest cruise port in Britain, hosting some of the largest ships in the world, doesn’t even monitor its air quality. Southampton City Council estimates that the port contributes up to 23 per cent of air pollution in the city, but the port either could not or would not verify this.

However, the Port of Venice’s head of environment, Marta Citron, did agree to talk to me. Every summer Venice makes the headlines as local protesters campaign against the huge cruise ships that pull into port, overwhelming the small city with people and pollutants. Marta was accommodating and friendly, and passionately believes that cruise ships are a positive for Venice as a whole. ‘It is the second-busiest port in the Mediterranean – the first is Barcelona,’ she tells me. ‘During 2016 we had more than 1,600,000 cruise passengers in Venice, and more or less 540 cruise ships … a voluntary agreement taken by the cruise companies decided to use in Venice only ships smaller than 96,000 tonnage, since 2014 … In 2007 the ships committed themselves to use 2.5 per cent sulphur content fuel from the entrance in the lagoon while at that time the MARPOL convention [required maximum content] was 4.5 per cent. Nowadays it is foreseen that all cruise companies that have signed the agreement commit themselves to use 0.1 per cent sulphur content in the fuel [in Venice] but during navigation they can [still] use 1.5 per cent … we have calculated it could be as much as 90 per cent sulphur reduction.’ I ask her about the protests of a few months previously, when nearly 20,000 Venetians voted in an unofficial referendum, with 99 per cent backing a motion to keep cruise ships away. ‘Erm, we don’t like to speak about this kind of things,’ she laughs, nervously, ‘it is a matter of communication more than environmental aspects, let’s say. We are technicians … the air quality is not a problem of Venice, it is a regional problem. In all the Venice region we have problems with PM, it is a continuous problem [in] relation to geographic and meteorological conditions.’ She tells me that PM levels are higher in winter, when the cruise ships don’t come, ‘so it is clear that it could not be related to ships’.

When I put this point to Sönke Diesener at NABU, he is brusquely dismissive of this view: ‘shipping contributes almost 100 per cent of the air pollution in Venice. The city has no cars and almost no industry … the central mode of transport is shipping. But of course, you have to ask what kind of shipping and which kind of air pollution we are talking about. I guess the answer they gave to you was only [based] on international shipping and is maybe linked to PM10. For all other air pollutants the ratio is way higher … We conducted measurements of ultrafine particles in Venice and the background level was around 5,000 to 10,000 particles per cubic centimetre. After arrival of a cruise ship it peaked at 360,000 and was on average above 60,000 … The port city of Hamburg just published its most recent air quality report. In the city of almost two million residents, lots of cars and industry, ships account for 39 per cent of NOx emission. There is no doubt that this ratio must be much higher in a city like Venice.’ As for higher levels in winter, PM10 is higher in most parts of Europe due to domestic wood fires and a lower atmospheric boundary layer. Northern Italy in winter is one of the PM10 hotspots of Europe. But nanoparticles and particle number are always higher close to the source – and in the case of port cities, that means shipping.

In many transport sectors, diesel is rapidly trying to clean up its act. Since September 2017, all new cars meeting the Euro 6 standard must also undertake a real world test on public roads – the emissions cap will be higher than in the lab tests, but it aims to bring them closer in line. In shipping, HFO will be banned by the International Maritime Organization from 2020, and fuel sulphur content will drop from 3.5 per cent to 0.5 per cent – albeit still a much poorer fuel than the diesel burned in road vehicles, which is typically lower than 0.005 per cent sulphur. But from the point of view of local air quality and nanoparticle emissions, there is simply no such thing as ‘clean diesel’. Diesel is a fossil fuel with a very high mineral and carbon content: by definition, you can’t burn it without health consequences for those that breathe it in. Forget about sulphur for a second – that is an issue that should have already been consigned to the history books, were it not for shipping literally burning tarmac in their engines. Even ultra-low-sulphur diesel fuel that meets the latest Euro 6 (or the equivalent US or Indian standards, EPA LEV III or Bharat 6) is terribly bad for us to breathe in. As Professor Paulson at UCLA describes, ‘when you live close to a roadway you have higher exposures to ultrafines … and you also have more directly emitted gases from all the little engines burning fuel … ultrafines are uniquely able to go places in one’s body that other particles cannot.’

Studies have found that nanoparticles smaller than 300nm (PM0.3) account for over 90 per cent of the total numbers of all particles. Remember the golf balls and the footballs? If you have the choice, it’s better to breathe in a few PM10 footballs with lower surface area than thousands of nanoparticle golf balls with all that surface area to do toxic damage to your lungs and with the unique ability to get into your arteries.

To avoid exposure to those, you need to avoid exhaust fumes. An experiment called DAPPLE (Dispersion of Air Pollution and Penetration into the Local Environment) highlighted a noticeable difference in the concentrations of nanoparticles between roadside and building-side locations in central London from 33,162 particles per cubic centimetre (cm3) up to 163,110cm3. Given that sides of buildings in central London are rarely very far away from a road, that’s a big difference within a matter of a few metres. Professor Prashant Kumar, of the University of Surrey, found that the number of particles by roadsides (327,000 cm3) in Delhi was ten times greater than background levels across the city (33,000cm3)4. Other studies of major cities have found that the particle count decreases by up to 40 per cent of their kerbside level within a distance of 10 metres (33 feet), after which they bunch together to form footballs, or evaporate.

The drivers sitting in diesel vehicles are exposed to the worst doses of all. At the British Heart Foundation Centre of Research Excellence in Edinburgh, David Newby recalls that, ‘The Sunday Times ran a mischievous article following [one of my studies] saying you shouldn’t cycle in the city. In actual fact, if you take a particle monitor inside a car, the levels are often three times higher than outside. Often there are no cabin filters in the air circulation. The air inlets at the front of the car are usually taking in the exhausts of the car in front of you.’ The closer you are to the road, he says, ‘the more pollution you breathe’, and the car driver is obviously the closest of all. As for cyclists and pedestrians, ‘the difference is exponential; you only need to be removed by one or two metres. Proper segregated cycle lanes that are only a metre away from the road can be enough.’ Jim Mills, with his experience servicing the AURN (Automatic Urban and Rural Network), backs this up: ‘it’s a well-known fact that if you’re driving through [a major road] you will get a higher dose than a cyclist or a pedestrian most of the time. Because you’re sitting in a bubble, you’re not getting the benefit of the air moving around you and diluting whatever pollution is around, you are essentially breathing air which is the exhaust of the car in front. We did some work for Dispatches on [UK] Channel 4 where we measured inside vehicles, on a bike, on a pedestrian and in a bus, and the highest levels in every one of those studies that we did was the driver. So, the guys who are producing the pollution are actually getting the highest dose … most people believe when they go inside their car and shut the door that the pollution is outside and they’re safe, but nothing could be further from the truth.’ The average adult in the US spends 55 minutes each day driving or being driven inside one such bubble of pollution.

Yet, in the race to reduce PM2.5 and NOx emissions from diesel car engines, the nanoparticle count and NO2 percentage may have gone in the other direction. At Millbrook, Phil Stones tells me that as the after-treatment of car emissions has become more and more advanced, ‘you start to create more primary NO2. So [in] older vehicles, most of the NOx would have been NO. Now, most of it will be NO2. In real terms, the NOx has gone down … while the proportion [of NO2] may have been 5 per cent, it is now 50 per cent … Total NOx has definitely gone down, but I don’t think total NO2 has.’ One of the most common after-treatments is urea injection, which turns the NOx into ammonia and then, in theory, turns the ammonia into water and (inert) nitrogen N2: ‘after-burning [treatment] is really expensive – for heavy duty [vehicles] especially. It has quite a lot of intelligence to it – there are a lot of sensors that monitor and feed back, it doesn’t just pump urea in because it feels like it. If you pump too much in, or when the exhaust isn’t hot enough, you create ammonia [NH3]. It’s called ‘ammonia slip’ – there’s a legal limit on the amount of ammonia you can slip. And of course you consume a lot, which has a disadvantage to the user because it costs money to put in, and inconvenience … if its after-treatment isn’t working right, it will give a lot of NOx output.’ By comparison, when a petrol engine ‘after treatment isn’t working right, it will give out hydrocarbons and CO [carbon monoxide]. So therefore the cold implications from petrol aren’t quite as bad in terms of NOx as a diesel.’

The particle filters fitted to modern diesel cars are also a double-edged sword. As the Euro standards for PM from diesel have gone down from 140mg/km in Euro 1 to 4.5mg/km in Euro 5 and 6, the only way to implement this was to fit filters that trap the particles. Particle traps are very effective in controlling the solid primary particles, including black carbon, with filtration efficiencies found to exceed 90 per cent. However, they don’t affect the secondary aerosol particles that form in the air after emission. And they do nothing for ultrafine particles. In fact, they can create more of them.

To measure particle numbers at Millbrook, Phil and his team take a sample by ‘attaching butanol to each particle and using a laser to count them’, he tells me. ‘That’s anything from 23 nanometres upwards … Gasoline direct injection [GDI] starts to produce more particulate number, compared to a non-direct-injection gasoline. Basically, because you’ve got higher compression and higher injection pressures … particulate number is regulated on diesels and petrol GDI and has been since Euro 5b on diesels and Euro 6 on GDI, down to 23 nanometres. Diesels have a high particulate number due to diesel particulate filters, which essentially take the big particles and burn them off, but this potentially creates a lot of small ones. GDI can create potentially more than a diesel particulate filter vehicle as the pressure of combustion creates lots of small particles. People are looking at gasoline particulate filters to reduce this. Particulate number has tended to go up since diesel particulate filters came in as the particles got smaller, but more of them.’ So far, he says, ‘there has been no regulatory reduction in the amount of particulate number that a vehicle can emit, but I would guess this is likely for Euro 7 if I was a betting man, or they will increase the range down to 10-nanometre particles to capture more.’

Particle numbers are also higher in winter in cold countries, because more nanoparticles are formed at lower temperatures and are less likely to evaporate. In 2016, a work group report on ultrafine particles by the American Academy of Allergy, Asthma & Immunology reported that ‘although improved engine and fuel technologies have significantly reduced the emission of particulate soot, [nanoparticles] can still be formed from vapor condensation and they can be even smaller than the emission particles’. Furthermore, the ‘introduction of catalytic converters … had the unintended consequence of shifting the bulk of the particle size distribution of exhaust PM to smaller diameters of 20 to 30 nm. Particle mass decreases with catalytic conversion, but the number of particles in the [nanoparticle] range increases.’5

It has been a huge own goal. Because of, not despite, all the cost and sophistication being thrown at the modern combustion engine, it is creating more nanoparticles for us to breathe. And it is doing so on the roads where we live and work, between ground and head height. Trucks are even worse. In a roadway tunnel study, trucks and heavy goods vehicles were found to emit 24 times more fine particles, and 15–20 times the number of particles, per unit of fuel burned, compared to light duty vehicles. Numerous studies going back to the late twentieth century have now concluded that whenever you’re near traffic, it’s nanoparticles that dominate the air; with each metre you step away from the road, the particle number count declines.

It reminds me again of that line from Professor Peter Brimblecombe, that environmental crises often repeat themselves, caused by rapid population growth, urbanisation, and fuel shortages followed by a new fuel more polluting than its predecessor. Modern engine technology is solving the problems of PM2.5, just as 20 years ago they tackled PM10. But they have made ultrafine nanoparticles and the total number of particles even worse. And these are the ones that we now know can get into the bloodstream and cause the most health damage. At Millbrook, Phil tells me that some cars are getting to the stage that half the cost of the car covers the after-treatment systems fitted to meet emissions standards; and yet, they still aren’t working to improve our health as they were designed to, and may in fact be doing the opposite (although this is not to say that old cars are somehow better – go back a decade or two and you may be breathing fewer nanoparticles, but with an added serving of SO2, higher NOx, higher carbon monoxide and much more black carbon). It all starts to add up to the conclusion: does our entire approach of burning fossil fuels for transport, within the streets that we live in, need replacing?

Notes

* Diesel is good for uses where you want high torque and low speed, such as tractors.

† We tend to think of European environmental regulations being stricter than the US, but California’s emissions regulations – for reasons I look at in detail in Chapter 7 – have consistently been a step ahead of the Euro standards.

‡ To further rub salt into the wound, it later transpired that VW developed the defeat device using a €400m loan from the European Investment Bank: the loan was supposed to help the carmaker develop an engine that would comply with increasingly stringent emissions standards. Which, I guess, they kind of did.

§ And it wasn’t just the cars. A spot check of HGVs in the UK between August and November 2017 found that 7.8 per cent of the lorries carried some sort of emissions cheat device. Out of 4,709 lorries spot-checked, 327 had been doctored to switch off emission controls. If the same proportion holds true across the country then around 35,000 illegally polluting trucks are operating across the UK.

‖ Which is equivalent to the amount of beer that Germany drinks in a month, according to 24coaches.com. Apropos of very little, but come on, how could I leave out a stat like that?

CHAPTERSIX

Struggling to Breathe

‘Every parent who comes to me now is talking about [air pollution]’, Dr Ankit Parakh tells me in his consulting room. A pulmonologist in paediatric medicine at BLK Super Speciality Hospital, he is on the front line of Delhi’s smog epidemic. The BLK is one of several huge, privately owned hospitals that sprang up since the Indian economy liberalised in 1991 and, like so much in Delhi, the divide between the haves and the have-nots is stark. Beggars, traders and rickshaws crowd around the hospital entrance and spill out into the road. Inside, the paediatric wing is full of kids with shiny Nike trainers and concerned parents tapping impatiently on smartphones. Framed photos of happy (and exclusively white) child models adorn the walls. Most patients are here for breathing and respiratory disorders. Approximately one in three adults in Delhi, and two in three children, have respiratory symptoms due to poor air quality.

‘In India the main problem starts around September, October – it peaks in November, then there is a down phase from December to February, and again peaks around March, April and May,’ says Dr Parakh. As we speak it is currently the November peak. ‘Anywhere in these seasons sees a surge of asthma and wheezing patients.’ Wheezing is a technical term. Asthma tends not to be diagnosed until around the age of five, because the symptoms in young children can resemble a wide range of other respiratory problems. Asthma-like symptoms therefore tend to be called ‘pre-school wheeze’, which sometimes develops into asthma, sometimes not. ‘Children who are exposed to air pollution, even pregnant mothers … the risk of these children actually having wheezing or asthma is definitely there,’ says Dr Parakh. ‘It is also related to the birth weight of the child … underweight children will have more wheezing episodes, the episodes are more severe, they are more difficult to control. And now it has been shown [air pollution] is not just a trigger that increases an underlying asthma, it is also an inducer which actually can generate an asthma … And it is not just about respiratory issues, people have shown hypertension.’ He pauses, and says again, eyeing the queue waiting outside, ‘There is a lot of concern amongst parents.’

I don’t keep Dr Parakh long, aware that his consulting hours are short. When I leave, however, he insists on showing me the way to the Metro station. I expect him to point out the direction as we reach the main BLK entrance, but in fact he walks me out and onto the road, into the throng of people and traffic that seemed so distant in the sterility of his consulting room. Crossing the road, his stethoscope and brilliant white coat wrap around him like a protective layer. No one dares bother him. A girl of maybe eight or nine is begging nearby. She is covered in dust, her hair matted to the thickness of a door mat. Every minute of her work and life takes place on Delhi’s roads, breathing in some of the very worst air in the world. Dr Parakh shakes my hand and ushers me protectively towards the station’s barriers, and I soon step into the clean carriages of a Metro train.

It’s easy to take everything Dr Parakh said as a given, but the definitive link between modern air pollution and health has only been made relatively recently. It had long been suspected. In the 1950s epidemiologists in California such as John Goldsmith made the link between air pollution and heart attacks, but the causal link was elusive and other factors such as high smoking rates were hard to separate. The International Agency for Research on Cancer suspected that PM2.5 was carcinogenic to humans in 1988, but only definitively said so as recently as 2013. Professor Bert Brunekreef, chair of the European Respiratory Society Task Force on Air Pollution, admitted in 2016 that ‘when I started my career in environmental health some 35 years ago, air pollution in western Europe was not seen as much of a public health problem.’

By the 2000s and 2010s, however, studies began coming thick and fast, showing that common pollutants in the air affect our health at every stage of life, in the womb, continuing through childhood, adolescence and adulthood into old age. Professor Brunekreef describes this as the ‘life course’ of air pollution on human health. And it is this ‘life course’ that I’m going to take you through in this chapter, from conception through to your likely premature death. Sorry, this might not be the cheeriest chapter to read. But boy, is it full of counter-arguments to the statement that ‘air pollution never did me any harm’.

In fact, it affects us before we are even conceived. The incidence of infertility has been creeping up in industrial countries from 7–8 per cent in 1960 to 20–35 per cent by the mid-2010s. During those 50 years, sperm concentration (the amount of sperm per millilitre of semen) almost halved. The sperm count in American males is decreasing by 1.5 per cent each year, and you don’t need to be a mathematician to work out that is unsustainable. Recent studies have started to suggest a strong link between ambient air pollution and this decline in fertility. Compounds of lead, cadmium, and mercury have long been known to damage the male reproductive system. Now smoke-derived PM2.5 and polycyclic aromatic hydrocarbons (PAHs) have been found to impair or disrupt sperm production too, even leading to DNA fragmentation in sperm. A long-term study in Taiwan of thousands of young adult males between 2001 and 2014, found that for every increment of 5mg/m3 of PM2.5 levels there was a decrease of 1.29 per cent in normal sperm morphology (the size and shape of sperm in a sample). An Italian study in 2003 found that tollgate workers – a cadre of society unusually exposed to traffic pollution – had significantly poorer sperm quality that other men from the same region.

If we do manage to conceive, then air pollution can damage the health of the foetus and lead to birth abnormalities. A study by Queen Mary University, London, in 2018, found that inhaled black carbon particles passed from pregnant women’s lungs to eventually cause small black spots in the placenta. A study in Ohio in 2017 of women living with high PM2.5 levels in the month prior to conception showed they had a higher chance of having a baby with birth defects – the most common being cleft lip/palate or abdominal wall defects –

compared to those that did not. In Wuhan, one of the most polluted cities in China, researchers also looked at all 105,988 births delivered in that city between 2011 and 2013. When these births were overlaid with the carbon monoxide (CO), nitrogen dioxide (NO2), sulphur dioxide (SO2) and ozone (O3) readings, the risk of babies being born with congenital heart defects was higher among women with greater exposures. Previous studies in California and Australia also found a higher risk of artery and heart valve defects from increasing O3 exposure during the second month of pregnancy.

High air pollution exposure also increases the likelihood of having a premature birth. A team of researchers from Stockholm, London and Colorado concluded in 2017 that as many as 3.4 million premature births across 183 countries could be associated with PM2.5, with sub-Saharan Africa, North Africa and South and East Asia most affected. India alone accounted for about one million avoidable premature births. A US study in 2015 found that just over 3 per cent (or 15,808) of pre-term births nationally per year could be attributed to PM2.5.

Smoking has long been known to produce underweight babies, so it’s no big surprise that traffic smoke does much the same. A London study in 2017 found that among half a million newborns, high PM2.5 levels were associated with a 2–6 per cent increased risk of low birth weight.*

Even in Sweden, with its comparatively clean Arctic air, newborns with relatively high exposure to traffic-derived NOx were consistently found to have smaller foetal growth in late pregnancy: for every 10mg/m3 increment of NOx, birth weight reduced by 9 grams. Most worryingly, a 2017 US study using six years’ worth of data covering nearly a quarter of a million deliveries found ozone was associated with a significantly increased risk of stillbirth. Both long-term low-level and short-term high-level O3 exposure consistently increased the stillbirth risk, leading the researchers to estimate that approximately 8,000 stillbirths per year in the US could be the result of O3 exposure.

For those who have a successful birth, air pollution raises the risk of pneumonia among under-fives and lifelong lung conditions such as asthma. According to the WHO, 570,000 children under five die each year across the world from respiratory infections such as pneumonia, while up to 14 per cent of children aged over five currently report asthma symptoms, almost half of them related to air pollution.†

The Columbia Center for Children’s Environmental Health goes so far as to say that air pollution is the root cause of much of the ill health in childhood today. Young children breathe in more air than adults relative to body weight, meaning they are disproportionately affected by air pollutants compared to adults. Babies under the age of one tend to breathe 600 litres per kilo of body weight, per day. By the age of four, as we grow, it reduces to 450 litres; by age 12, it’s 300 litres; and by the age of 24 it plateaus at 200 litres per kilo per day and stays there for the rest of adulthood. When exposed to a polluted environment, children therefore suffer the ill effects three times more acutely than adults. Babies’ immune systems are also not fully developed and are more vulnerable to infections and almost defenceless against toxic exposure. Children are the first and worst victims of lead pollution too, because their body’s immaturity makes them most susceptible to neurological injury, leading to lowered IQs, reading and learning disabilities, impaired hearing and behavioural problems including ADHD.

There remains, however, a question of causality around much health evidence regarding air pollution. By their very nature the studies are based on the epidemiology – i.e. health trends across a population. You can’t put 100 kids in a lab, expose them to pollutants, then cut them open to see what happened. So epidemiological evidence will always suffer from the kind of argument that goes, ‘just because 30 per cent of the population gets cancer and 30 per cent of the population eats cornflakes for breakfast, does not mean that cornflakes cause cancer’. When an epidemiology study is repeated in different places, and consistently comes up with the same results, however, it becomes very hard to ignore. And the epide­miological evidence for air pollution reducing the size of children’s lungs is, I would argue, the most compelling example.

The Californian Children’s Health Study is one of the most comprehensive investigations of the long-term consequences of air pollution that we have. Starting in 1993, more than 11,000 schoolchildren were selected from 16 communities, and their lung function measured annually while air pollution levels were measured continuously. The performance of the kids in forced expiratory volume or FEV tests (how much air a person can force out in one second) declined according to exposure to NO2 and PM2.5. The proportion of 18-year-olds with a low FEV was four times higher in the communities with the highest PM2.5 levels compared to the lowest. Their lung growth had actually been stunted. Living within half a kilometre of a freeway was associated with a 2 per cent deficit in forced vital capacity (the total amount of air in the lungs). International study after international study has since found the same, including in Mexico, Austria, Norway, Sweden, the UK, and the European Study Cohorts for Air Pollution Effects. A three-year study in China found that an increase of 10mg/m3 of PM2.5 was associated with a loss of 3.5ml in FEV capacity. In Delhi in 2012, one in three children in the city were found to have reduced lung function.

Professor Chris Griffiths, from the Centre for Primary Care and Public Health at St Barts Hospital and a working GP, was involved in a six-year children’s lung capacity study in London which concluded in the 2010s. The children living in areas with high levels of particulates and nitrogen dioxide had their lung capacity reduced by up to 10 per cent. I asked him if there remained a ‘causality problem’: ‘You’re not going to get a randomised clinical trial of air quality interventions, it doesn’t work like that,’ he argues. ‘You’ve got to reach a point where you say “Well how strong is the evidence? What’s the quality? What’s the causal inference?”’ Regarding his own lung study, he says ‘the mechanisms that underlie these observations are not clear, but that doesn’t mean that those associations aren’t there or aren’t important, just because we don’t quite know how air quality mechanistically reduces lung growth. But because the data is quite consistent wherever these studies have been done, Europe, Scandinavia, Boston, California and then most recently London, studies in different settings with different pollutants, yet they appear to be saying similar things about lung growth.’ The conclusion drawn by many, including the UK’s Royal College of Physicians, is that there is little reason to ‘doubt that air pollution adversely affects the normal growth of lung function during childhood, right up to the late teens’.

The ongoing Californian Children’s Health Study, led by William ‘Jim’ Gauderman, a professor of preventive medicine at Keck School of Medicine USC, also continues to add to the weight of evidence. While in 1993 Los Angeles had some of the worst atmospheric pollution in the world, by the 2010s it was still bad but much improved. His recent papers have been able to flip the arguments against epidemiology on their head. Cornflakes don’t cause cancer. But if you reduce the consumption of cornflakes‡

and the prevalence of cancer drops at precisely the same rate, then perhaps you do have to question what you’re putting in your bowl each morning. Comparing three cohorts of the Californian Children’s Health Study from 1994 to 2010, Gauderman and his team found that the mean four-year growth in FEV (volume of breath in one second) increased by 91ml for every decrease of 14ppb in NO2, and similarly for PM2.5:1 in other words, whenever the nitrogen dioxide and PM2.5 levels went down, the children’s lung function showed a marked improvement.

As our life course of air pollution moves from childhood and into adolescence, low FEV comes with an increased risk of cardiovascular disease – given that you’re missing 20 per cent of your possible lung capacity, you’re less able to exert yourself during exercise. Equally logical is the fact that asthma is both caused by high pollution and irritated by high pollution: ozone, nitrogen dioxide and PM2.5 all cause inflammation of the airways, and airway hyper-sensitivity is the defining characteristic of asthma. The UK Chief Medical Officer’s report suggests a link between diesel particles and the high asthma rate of 1 in 12 adults and 1 in 11 children in the UK. But respiratory conditions are only the most obvious ones affecting young adults. Our brain capacity may also be reduced upon entering adulthood. European studies have associated traffic pollution with lower cognitive development in junior school children followed by sustained attention deficit in adolescents. Studies in Mexico City have also revealed elevated levels of inflammation in the brains of children exposed to high air pollution, resulting in cognitive deficits.

If you grew up in a rural oasis of clean air and avoided all the health problems thus far, then if you move to a city (or your village turns into a city, as several have in Asia in recent decades) at any point in adulthood, plenty of health problems still await. Sticking with the brain for the moment – and let’s face it, brain damage is one of our biggest fears – a Chinese study exposed lab mice to NO2 inhalation and found it caused deterioration of spatial learning and memory. The study authors memorably describe air pollution as ‘a multifaceted toxic chemical mixture capable of assaulting the central nervous system’. Outside of the mice fraternity, numerous epidemiological studies have linked NO2 pollution to an increased human risk of neurological disorders, including reduced cognitive and attention scores. One fascinatingly precise study of white matter loss among elderly women, made using MRI (magnetic resonance imaging) scans, suggested that for every 3mg/m3 increase in PM2.5, white matter loss increased by 1 per cent. Suicidal depression has even been suggested by more than one study to increase one to three days after a peak in PM2.5 and NO2.

For those who trust their gut more than their brain, a team from UCLA found that exposure to air pollution changes the composition of our gut bacteria. This leads to a whole range of health problems, including the circulation and build-up of cholesterol in the bloodstream. Other studies have found an association between air pollution and intestinal disease, appendicitis and even digestive tract cancers. There’s a direct link for the chunkier PM10 here too: while large enough to be ejected from our throat and lungs via mucociliary clearance (our respiratory system’s first line of defence, whereby a layer of fluid and mucus is constantly being propelled upwards for us to spit out), anything on the surface of the PM10 can be dissolved by saliva and find its way down into our gut. Depending on the chemical nasties on the surface of these coarse particles, this can lead to an imbalance in gut bacteria, or cause the chronic inflammation that leads to appendicitis or cancer.

Many airborne pollutants are known carcinogens. Polycyclic aromatic hydrocarbons (PAHs), for example, have toxic effects which can cause cell damage, leading to mutations and tumours. Long-term occupational studies of workers exposed to PAHs have shown an increased risk of skin, lung, bladder and gastrointestinal cancers. The PAH known as benzo[a]pyrene, emitted in high abundance by stubble burning, was named as a human carcinogen as early as the 1980s by both

the International Agency for Research on Cancer (IARC) and the US EPA. Since then the EPA has classified other PAH compounds as carcinogenic, all with confusingly complex names such as benz(a)anthracene, benzo(b)fluoranthene and indeno(1,2,3-cd)pyrene.

In summary, pick any major organ or body part and there will be a disease or defect related to air pollution. How about breast cancer? In Hong Kong, a 10ug/m3 increase annual PM2.5 exposure was found to increase the risk of breast cancer by an astonishing 80 per cent. Did someone say kidneys? Research from St Louis, Missouri, looked at more than eight years of data from nearly 2.5 million military veterans, and found that the veterans’ kidney function worsened over time according to the level of pollution they were exposed to. Higher PM concentrations were associated with an increased risk of end-stage renal disease, after which a person requires kidney dialysis to stay alive.

Air pollution can even change how our DNA behaves. Genes – the segment of DNA that tell the cells of the body what to do, and when – are controlled by a chemical switch known as a methyl group. These methyl groups can, in effect, switch a gene on or off. A 2014 study by the University of British Columbia put 16 volunteers into an enclosed booth for two hours, giving half the participants clean air to breathe while the other half breathed diesel fumes equivalent to a busy highway. The methyl groups changed at about 2,800 different points on the DNA of people who breathed in diesel fumes, affecting about 400 genes. No similar changes were seen among the group breathing clean air. Until that experiment, scientists mostly thought that DNA responded primarily to long-term exposures. A similar Chinese study in 2017 compared traffic police officers to office-based police officers and found that DNA damage was significantly increased among the traffic cops compared to the pen-pushers at City Hall.

But it’s the effect on our cardiovascular system that is most fatal across the adult population. Yes, even more so than cancer or lung disease. Air pollution causes thinning arteries, blood clots, heart attacks and strokes. As fine nanoparticles enter the bloodstream through the walls of the lungs, they cause increased inflammation, resulting in changes in heart rate, heart rhythm and blood pressure. This is not just from chronic, long-term exposure, but also from short-term. In Beijing, data from daily cardiovascular emergency room visits were collected from ten large hospitals for the whole of 2013, the year of the now-infamous Airpocalypse. A 10mg/m3 increase in PM2.5 was associated with a 0.14 per cent increase in cardiovascular emergencies per day. That may not sound like much, but given the monthly high and low of PM2.5mg/m3 in Beijing can differ by as much as 300mg/m3, that could cause an increase in cardiovascular emergencies of 4 per cent, a significant burden on health services.

Years before David Newby’s gold nanoparticles study (see Chapter 3), his first exposure chamber study exposed volunteers to street levels of air pollution and found that blood vessels were more likely to clot compared to being exposed to clean air. ‘I like to call it a vascular stress test,’ he says, ‘What we do is put a little needle in the artery in the arm, and through that we infuse some agents to stimulate the blood vessels to relax and dilate. What we were able to show is when you are exposed to dilute [vehicular] exhaust, your blood vessels don’t relax as much … and that means blood flow might be slower. We also tested a protein that is released from some of these cells, called TPA, which stops clots forming within the blood vessel, so the blood continues to flow … It is a clever way the body is able to both continue blood flow but also prevent you from bleeding to death. What we found is the release of this TPA is lower when exposed to air pollution than if not. So, the defence mechanisms are hindered.’

Newby’s team took this research a stage further, taking blood from a human volunteer and passing it through an artificial coronary artery. Inside this ‘artery’ was a strip of pig aorta – one of the heart’s main pumps, at the top of the left ventricle – obtained from an abattoir. By cutting some of the surface off the pig aorta, it modelled what might happen when you have a heart attack and part of the artery bursts, exposing the deep layers of the artery.§

‘What we found was that when you expose people to dilute diesel exhaust, the amount of clot formed on that strip increased. And when we put a filter in and took the particles out and redid the test again, the amount of clot came down to normal levels. So, it does seem that the diesel exhaust exposure makes the blood thicker. So that’s three effects we’ve got: one, the blood vessels don’t relax as much; second, they don’t release as much of this clot-dissolving protein TPA; and when an artery is damaged, more clot forms. These are powerful mechanisms in the relationship between heart attacks and strokes.’ I ask if he also attempted this test with gas pollutants, without PM. ‘That’s right, we did some NO2 exposures, and also did it with ozone – just the gases on their own, no combustion-derived particles, and we didn’t see any effects. People have argued with us about that, they were expecting certainly NO2 to have an effect, but we didn’t see anything. Some people have argued that it is the combination of the NO2 with the particles that causes a problem, and that is possible.’ However, a team from University Hospital Jena, Germany, subsequently found a direct link with NO2 too. Their 2018 study of 693 heart attack patients found that a rise in NO2 of more than 20mg/m3 within 24 hours increased the risk of heart attack by up to 121 per cent, while a rapid hourly increase of NO2 of just 8mg/m3 increased the risk of heart attack by 73 per cent.

If we actually manage to reach old age, then air pollution severely undermines our quality of life. In a study of the elderly in the United States, PM2.5 and NO2 exposures were significantly associated with type 2 diabetes. Associations between air pollution and serum glucose, a measure used to assess diabetes status, have been reported even with short-term NO2 and PM2.5 exposure. The same mechanisms that affect the young developing brain also set to work on diminishing brain function towards the end of life. Experimental studies have shown that air pollutants cause neuro-inflammation, neuron damage and blood–brain barrier problems. One South Korean study of Parkinson’s disease hospital admissions from 2002 to 2013 found a ‘consistent and significant association’ between short-term exposure to air pollution (except for O3) and higher rates of Parkinson’s disease patients being admitted to hospital. The US Women’s Health Initiative Memory Study (WHIMS) also enrolled over a thousand older women with no previous signs of dementia between 1996 and 1998, and studied them over six to seven years with regular brain MRI scans. The women who lived in places with higher long-term levels of PM2.5 were found to have smaller brain volumes, not explained by demographic factors, socioeconomic status, lifestyle or other health characteristics. Frank Kelly’s research group at King’s even looked at 8 years worth of GP records of over 100,000 Londoners aged 50–79, and found that those living in areas with high NO2 and PM2.5 had a 40 per cent greater risk of developing dementia than those living with low pollution. The potential mechanism for this was suggested by work based in Lancaster, England (a relatively clean-air city, by UK standards). It found that local traffic pollution included 200 million metal nanoparticles per cubic metre, which it believed caused inflammation in the brain. I talked to Jim Mills, MD of Air Monitors, shortly after this study was released. ‘Can you imagine what the political change of opinion would be if that gets proven?’ he asked. ‘That all the problems we have with rising dementia in our population, which is one of the most scary things I think that we face at the moment, are actually down to our use or overuse of the internal combustion engine producing these tiny particles?’

Underlying this dementia research, and almost all the health problems throughout the life course of air pollution, is oxidative stress. Oxidation is happening all the time, within our bodies and throughout the natural world. Highly reactive compounds search for electrons to steal from others, which sets off a chain reaction as compounds either lose an electron or try to replace one. The easiest example to picture is rust. When iron meets oxygen it creates iron oxide, or rust – oxygen steals electrons from iron, which forms a new, weaker compound. This flow of electrons is necessary for life to exist, driving photosynthesis by oxidising water (H2O), which releases electrons to turn carbon dioxide (CO2) into carbohydrates and oxygen (O2). It’s literally why we breathe: oxygen comes in and reacts with big hydrocarbon molecules like sugar to break them down and give us energy. Almost everything eventually oxides down to CO2 or H2O, which we breathe out and pee out, and the process continues ad infinitum. So, oxidation equals good, right? Not completely. Oxidation is both caused by and causes free radicals, such hydroxyl (OH), those atmospheric firecrackers that for the milliseconds of their existence are desperately looking for a fight. Again, free radicals are natural and necessary, and form part of our immune system. It’s just that we don’t want to encounter an unnatural number of them. If we do, it causes more oxidation than our bodies can deal with: this is known as oxidative stress.

Professor Frank Kelly, chair of the Committee on the Medical Effects of Air Pollutants, took up his first lectureship in the late 1980s looking at the effect of the free radical OH on premature babies. ‘Because their lungs are under-developed they have to get extra oxygen in incubators to keep them alive and their brains functioning. But in giving their bodies more than the 21 per cent oxygen which we breathe normally – some of these kids require 90–100 per cent oxygen – it actually ends up damaging the tissue and they have a whole range of conditions that develop in their eyes, brains and lungs, because of this high oxygen concentration. And the reason this happens is because of free radicals.’ Kelly’s work in this then brand-new field of research found that while the lung tissue is protected from free radicals by the natural antioxidants in the body, if OH enters in high enough concentrations and for long enough, then the natural defences are overwhelmed – the atmospheric guard dog is unleashed, if you will, desperately snapping at soft body tissue and causing inflammation.‖

Our body is under constant oxidative stress, even without pollutants getting involved. It causes damage to cells, proteins, DNA, and may even be the whole reason why we age. ‘It’s a bit like if you leave butter out too long it goes rancid and oxidises,’ says David Newby. ‘There is a [theory] that atherosclerosis, this fatty deposit in your arteries and the underlying cause of heart attacks and strokes … gets oxidised [by pollutants], and it’s this oxidisation that causes [heart disease].’ The bad news is that NO2, pumped into every town and city in the world by car engines and boilers, happens to be a highly reactive free radical. Ozone, while not a free radical, is highly reactive and has an abundance of oxygen atoms, so tends to get into a fight with free radicals wherever it appears. PM and black carbon are coated in toxic material full of potential free radical stimulants. Our bodies’ defences are therefore overwhelmed by this multi-pronged attack. ‘Certainly when you look at the oxidative potential of [air pollution] particles, you really get a huge signal,’ says Newby. ‘In one of the studies I was telling you about – [blood vessels] not relaxing as much – part of the reason that we think they don’t relax is that the oxidative stress consumes the mediator which makes the arteries relax … before it can have an effect on the blood vessels.’

In 2017, another team from Edinburgh Napier University looked at the release of defence proteins and peptides, which effectively march out to defend the body at the first sign of trouble. One such peptide, known as LL-37, present in saliva, tears and lung fluids, has many immune system functions including guiding inflammatory cells to a wound or infection. The Napier team examined the effect of black carbon particles of 14nm (within the range of particles able to enter the bloodstream) on LL-37. Even with relatively low concen­trations of nanoparticles, the presence of LL-37 seemed to reduce; at high concentrations, it disappeared. On further investigation, it was found that the black carbon particles had grown in size. They had, like a cartoon snowball rolling down a hill, simply stuck the peptides onto their own surface, and in doing so rendered them powerless. The bundles of carbon particles and LL-37 no longer had any effect on the bacteria present in the test.2

To make matters worse, according to Frank Kelly, activated inflammatory cells also generate and release large quantities of free radicals themselves as a form of defence. In the absence of any invading organisms to kill, these free radicals can turn on their host and start attacking local cell tissue components. Many of these reactions happen first in the lung lining fluid, the first line of defence the pollutants encounter when we breathe them in. Chemical chaos ensues, and an inundated immune system resorts to calling the body’s last line of defence: a tank battalion of inflammatory cells. That chain of events precedes anything from an asthma attack to the early formation of tumours. Presumably, then, taking antioxidant supplements could counter this effect? ‘If you look at trials that give antioxidant vitamins, they have absolutely no benefit whatsoever,’ says David Newby, disappointingly. ‘It doesn’t matter how many antioxidants are flowing around in your bloodstream … that local burst is very difficult to prevent.’

The heart is particularly susceptible to oxidative stress, as it is an extremely active organ with a high metabolic rate and high energy demand. It’s not exposed to the air, so can’t oxidise and spoil in the same way as a lump of butter, but nanoparticles entering into the bloodstream can carry oxidative molecules on their surface. According to Newby’s gold study, once inhaled, nanoparticles of 30 nanometres (83.3 times smaller than PM2.5, most coming from car engines) can travel anywhere the blood goes, which is basically our entire body. They are reactive and toxic due to their large surface areas (remember it’s golf balls versus the footballs?), leading to oxidative stress and inflammation. The European ‘Exposure and Risk Assessment for Fine and Ultrafine Particles in Ambient Air’ study investigated the health effects of nanoparticles in coronary heart disease patients. It observed that the risk of developing ischemia – when blood flow to your heart is reduced, preventing it from receiving enough oxygen – was significantly greater two days after exposure to increased outdoor levels of nanoparticles. A US report on nanoparticles similarly concluded that they have ‘greater oxidant potential and were much more prone to introducing cellular injury compared with PM10 and PM2.5’.

Ken Donaldson, a recently retired particle toxicologist from the University of Edinburgh’s MRC Centre for Inflammation Research, spent the latter half of his career studying the toxicological impact of nanoparticles. I ask him if, in general, nanoparticles from combustion sources are more toxic than those from non-combustion sources. ‘Yes’, is the simple answer: ‘combustion-derived particles are the big hazard [because they] have more than just a larger surface area, they have metals and organics. Both of these can undergo redox cycling¶

to produce oxidative stress.’

Add all these health effects up, and what do you get? According to the WHO, the answer is 4.2 million premature deaths a year (or 7.4 per cent of all deaths) from outdoor air pollution. The WHO even provides exact figures for each country. In the Philippines in 2012, for example, the WHO attributes 28,696 deaths to ambient air pollution. But unlike during the great smogs of London and Donora, these aren’t numbers that represent people suddenly dropping down dead from air pollution. So how do we end up with such precise figures?

Frank Kelly, chair of COMEAP (Committee on the Medical Effects of Air Pollutants), tells me these figures are calculations based on the total years of life lost: in the UK, air pollution is shortening life expectancies by three to seven months on average, amounting to 340,000 years lost across the total population. Divide this by the average lifespan, and you get to a figure of around 40,000 ‘deaths’. Kelly even uses the word ‘guesstimation’. Given that almost every article on air pollution leads with these annual death figures, isn’t that a bit problematic? Numbers of deaths, says Kelly, is something that the public can easily understand, whereas ‘they don’t understand and don’t worry about losing three months off our lives. But even that is a generalisation – some people are losing a day and some are losing ten years … My response is, we talk about 89,000 premature deaths from smoking [in the UK] each year – Department of Health, official figures – that uses the same terminology, the same approach. It just helps to put it in terms of the risks we face in society. Breathing poor air in cities [kills more people than] drinking, more than obesity, way more than road accidents. But don’t take [the numbers of deaths] as being 100 per cent accurate, because we haven’t got that precision.’

COMEAP has regularly stressed that air pollution is shortening the lives of many more people than just those 40,000 a year in the UK. We will all die. But air pollution is likely to make you and me die sooner than we would have done otherwise. And, as it is related to many chronic, debilitating diseases, it is likely to make some of the years that we live much more painful than they would have been otherwise. The UK Chief Medical Officer’s annual report recommends ‘not just focusing on mortality, but using data that capture the full health consequence of pollution on morbidity, mental health impacts, and impacts on quality of life … Life-years (or quality-adjusted life years) are more appropriate for analysing policies than numbers of deaths, as it is when people die rather than whether they die that matters.’

Professor Chris Griffiths at St Barts Hospital believes the annual death figures aren’t entirely helpful: ‘there are other important ways of expressing the adverse health effects … I think the lung growth data is a little bit easier for people to get their head around, by saying “These children’s lungs aren’t as big as they should be [because of air pollution]”.’ I ask him if the focus should shift to quality of life. ‘Yes, yes. You’ve got more people with asthma, people with asthma having more asthma attacks, more pneumonia, more hospital admissions, more strokes, heart attacks, premature births, small birthweights, these are all statistically significant adverse health effects and they all add up, as you say, to impaired quality of life. I think we are getting stuck on the issue of mortality and air quality … the truth is if people’s lives leading up to that point of death are miserable because their lungs are destroyed … then air quality is important.’

Our life expectancy, then, is being reduced on average in Europe by 8.6 months to a year by PM2.5 pollution and in India by 1.1 to 3.4 years (or as much as 6.3 years in Delhi). But take away the source of the PM, and we see health dramatically improve. Frank Kelly writes of ‘consistent evidence that a reduction in the level of particulate pollution following a sustained intervention (mainly regulatory actions) is associated with improvements in public health’. The WHO argues that ‘health improvements can be expected to start almost immediately after a reduction in air pollution’. The deaths of many people every year are undoubtedly being caused by air pollution, although we’ll never have an exact number. But all of our lives are being shortened, or worsened, by air pollution. If we want it to stop, we have to fight back.

Notes

* The reasons include placental inflammation, impaired oxygen transport across the placenta, unstable blood pressure and even inflammation of the foetal lungs.

† The WHO also suggests that warming temperatures and ever-rising carbon dioxide levels may increase pollen levels, making asthma rates even worse.

‡ Other breakfast cereal analogies are available.

§ Of all the scientists I spoke to, or read about, for this book, I felt that David Newby pushed the boundaries more than most in the quest for the causal link to air pollution. I was surprised when, in early 2018, a scandal broke in Germany following the revelation that VW had taken part in an NO2 exposure test on 25 human volunteers. It led news bulletins not just nationally but internationally. Chancellor Angela Merkel called it ‘wrong and no way ethically justified’. Newby had been conducting similar studies for years. The German newspaper Tageszeitung probably put it best in saying that ‘human volunteers only had to inhale exhaust fumes for a few hours, people [in cities] … have been breathing in levels of nitrogen oxide far higher than EU limits for years’. Quite.

‖ Many animals are affected in the same way, too. Studies have found that European house sparrows, blue tits and great tits from areas with higher vehicle pollution had increased levels of oxidative stress compared to birds in less polluted areas, while street dogs culled in Mexico City were found to have far greater lung and brain inflammation than rural dogs.

¶ Redox is a portmanteau word combining ‘reduction’ and ‘oxidation’. As one molecule loses an electron (thanks to a scrap with a free radical) it is oxidised, while the molecule gaining an electron is reduced. I know, it sounds like it’s the wrong way round. Thanks scientists. The other way to think of it is the oxidation state of a molecule is either increased (oxidised) or decreased (reduction) – and so on, in an endless cycle.