#4.2 - Ben Jones: Unpacking the Science of Household Air Quality, Cooking Emissions, and the Impact of the Built Environment on Health

Ben Jones - Associate Professor at the University of Nottingham in the Department of Architecture and Built Environment.

With a  Master's Degree in Aeronautical Engineering, he worked as a Senior Software Engineer at BAE Systems before completing an Engineering Doctorate in Environmental Technologies at Brunel.

He was a Research Associate at University College London for two years before taking the post in Nottingham in 2013.

Ben's work focuses on measurement and modelling approaches to the indoor environment. He is particularly interested in the energy-efficient ventilation of buildings and its relationship with indoor air quality and occupant health.

Now and then, a piece of work comes along that has the attention of the room. And the work that Ben and his colleagues have been responsible for on Harm is right up there.

It's about the Harm that pollutants may cause and ways we can better define it and ultimately what we consider good or bad indoor air quality.

We talked about much more, including relative risk, cooking pollutants and what he is working on right now.

As always with Ben it was a genuinely fascinating conversation. I hope you enjoy it. Thanks for listening.

Ben Jones - LinkedIn - https://www.linkedin.com/in/benjamin-jones-0686a214/

Ben Jones  - Nottingham University - https://www.nottingham.ac.uk/engineering/departments/abe/people/benjamin.jones

A preliminary assessment of the health impacts of indoor air contaminants determined using the DALY metric - https://www.tandfonline.com/doi/full/10.1080/14733315.2023.2198800

AIVC - https://www.aivc.org/

ASHRAE 241 - https://www.ashrae.org/technical-resources/bookstore/ashrae-standard-241-control-of-infectious-aerosols

Chapters

• 0:00 Air Quality and Engineering in Practice
• 6:51 Handling Uncertainty in Modeling Airborne Transmission
• 22:08 Ventilation and Uncertainty in Air Quality
• 30:17 Cooker Hoods and Air Quality Regulations
• 35:57 Changing Views on Health and Well-Being

Transcript

Simon: 0:00

Welcome back to Air Quality Matters. This is my conversation with Ben Jones, part two. If I could just turn it to you a little bit to start with, how did you even end up in this field of Air Quality and engineering and the built environment? What was your pathway to being here and what are you guys doing at Nottingham now day to day from a work perspective? What are you currently looking at and working on?

Ben: 0:44

I ended up here by accident. It's a happy accident, though. I'm an aeronautical engineer by training and I spent five years in practice working for British Aerospace. After five years I was a senior software engineer for them, but after crunching code for five years, I realised that that was not something I wanted to do for the rest of my career. I founded the Doctorate with Ray Kirby and Maria Colicatrone at Brunel, which enabled me to work in industry and do research for industry while also studying sustainable development, which was brilliant. So I ended up with a company in Highwick and Monodraft I still have great links with.

Ben: 1:30

We worked out how their natural ventilation systems were performing in schools. So there was the labour schools for the future programme that was going on back then, every investment in natural ventilation. Two years then at UCL, where they changed my perspective from looking at technologies to the bigger picture stuff, and of course, at UCL, if you go to the main building, you find Jeremy Bentham in a box in the corner, and Jeremy Bentham was the guy who came up to his preserved body. Is there, believe it or not. He was the guy who came up with the phrase approximately having the greatest effect on the greatest number of people and, I would add, for the least number of dollars, sort of utilitarianism. And we were looking at stocks of homes and what happens when you decarbonise them and what the unintended health consequences. And we were using the quality at the time and one of the things we were looking at was radon, because as you tighten, improve air tightness in radon affected areas and you don't provide concurrent ventilation, you end up with what were previously medium risk homes becoming high risk homes and high risk homes becoming really, really dangerous homes. So we were trying to quantify that.

Ben: 2:57

I then moved to Nossi Human 2013,. I've been here for 10 years. When you first arrive in an academic post, you're desperately trying to set yourself up. But I took Max's plot my favourite nerdy plot and realised that I should be looking at particulate matter in homes. I managed to get hold of a couple of sensors, put it through some small budget wins, project wins, and started looking at emissions from cooking in student kitchens and I had one student measure emissions from 150 pieces of toast. He said he never wants to eat toast ever again.

Ben: 3:37

But that work was instructive because it told me there was huge variation in what we were witnessing.

Ben: 3:44

And then we went to work with TNO, with Walter Bosboom and his colleagues at TNO in the Leatherlands, and we started to cook dinners again and again in the same way and we observed consistent patterns but uncertainty in the emission rates. And the reason we were interested in the emission rates is because you generally want to model what's going on in the building. It's very time consuming, expensive and intrusive to go into people's homes and measure stuff, so we wanted to model what might be going on at stock level and so we needed good data to go in there and we needed data where we could quantify the variation that we were seeing in numbers. So that's how we got into it really, and later more funding became available to start looking at the harm from this rather than just comparing against World Health Organisation threshold values that we always knew to be a problem, so that now we're in a process of trying to say well, if we install a cooker hood in every home, what might be the effect on population health in, say, London or in the entire UK?

Simon: 5:09

Just goes to show good science doesn't operate in a vacuum, that when we're talking about particular matter from cooking, there's often a lot of work that's gone in into trying to understand what the emission rates of certain activities in those kitchens are, so that we can model outcomes more accurately.

Simon: 5:27

And you started to pick up the moniker of Mr Bernt Toast there, for a while I think through some of your studies and I've tried to explain to people that you read the papers of some of these studies that looking at something that seems very singular in approach of what is the emission rate of frying bacon or burning toast, and there's great descriptions in the papers of trying to have a repeatable method for turning bacon over and measuring the particulate matter. But that's the kind of science you have to go through for a number to appear in a formula. I guess that you can stand over and have some level of confidence that a combination of cooking breakfast and cooking lunch and using the oven and frying gives you these types of emission rates. That's the kind of the legwork that has to go in behind those numbers. Yeah.

Ben: 6:18

Yeah, so I had a completely different question, which was a population question. I had a model to do that and then didn't have the data. So I've got to get the data and that's where it took me I'd rather. Personally, I'm more interested in models and statistics and playing around with code back to my British aerospace days, because measuring stuff is so hard so there's too many confounding issues to it. So, but nevertheless, that's what we did.

Simon: 6:51

And what are the kind of things you're doing at the moment? Where's your focus? I mean, obviously, I imagine this harm work has been a real area focus for you and your team now for a while. Are you kind of nosing around in other directions yet? Is there other stuff that's kind of pushing your buttons?

Ben: 7:10

Yeah. So I think I think we're now trying to evangelize and trying to try to see if people think there are problems with this method, the problems with our approach, and we get asked the same questions wherever we go. So we're getting very good at answering them and we've covered a lot of them today, and so we're now just trying to work out what are the barriers to actually getting this into regulation. How can we, how can we, persuade different regulators to do something about it? So we've already got a draft addendum to Ashway 62.2, which is very exciting, but it doesn't necessarily mean it will be accepted. We've still got a lot of work to do, to do there.

Ben: 7:52

Alongside I mean, curiously, you know the beginning of the pandemic when it became evident that there was some form of airborne transmission of SARS-CoV-2, I suddenly thought well, you know, it's contained in aerosols.

Ben: 8:10

My toaster missions behave in a very similar way to these, these, these aerosols. There are a couple of other things that change quite notably, but the essential physics in a space doesn't change very much. And there was a terrific research project run by Malcolm Cook at Loughborough called AirBods, and I invite everybody to go and have a look at airbodsorg to see the outputs from that project and I got to work with Chris Iden, who I've known for a very, very long time and if there was ever a skill set ready for a pandemic it was Chris's skill set. So Chris was a PhD in biochemistry. He then worked for 10, 15 years in the ventilation industry, so he's got all this knowledge of viruses, which is what his PhD was on, and he did a post-op too and he knows how ventilation systems work in buildings. And we started to get together to look at some risk models which were incorporated in the few sage documents. They were the foundation of much of the SIPC advice and guidance given during the pandemic and have recently underpinned the development of AASHROY.

Ben: 9:26

Standard 241, Control of Infectious Aerosols, which is the first standard of its kind, I believe.

Simon: 9:33

That's really interesting because it was something I wanted to ask you about. I mean, people have to, I guess, try and cast their minds back. What now? We two and a half, three years even, where we started to understand that this thing was a serious risk. We wanted to develop tools to understand how much of a risk this was and what mitigation we could put in place to reduce that to a certain level. And hey, presto, up pops the maths. We have to start running models and scenarios to figure out what happens if one person is infectious versus three. What happens if the ventilation rate is X versus Y? What's the impact does the size of a room have?

Simon: 10:23

All of these questions that people are never logically want to ask need to be run through various scenarios. So there you start to get into things like relative risk, which is an area I know you've spent quite a bit of time working in, and my apologies if I completely butcher something I remember you saying, but it was something along the lines of having done a lot of that work, that you could say with 100 percent certainty that the chances of getting COVID right in a model was somewhere between zero and 100 percent. I don't know if I got that quite right, but I think the overwhelming output from that is that there's so many variables, there's so many, so much relative risk and uncertainties in the real world. How do you start to model that? So my question to you is taking us back to that and what was driving some of those questions. And when we work in the real world and a world of uncertainties, how do we practically handle that from a scientific perspective or an engineering perspective? How does that translate into outputs like two for one, for example?

Ben: 11:38

OK, there's a lot to unpack there.

Simon: 11:40

I know that's a big question, apologies.

Ben: 11:42

So one of the things I was hinting at when we were talking about the emission rates from toast and plates of dinner that we were doing was that you don't get the same answer every single time. It's what we call uncertainty in it, and if you think about this, you know with your own body and at all are you signing?

Simon: 12:03

I'm six foot eight, Ben, you know that About five eleven. Why aren't you in basketball?

Ben: 12:09

So I'm 186 centimetres. The average person in the UK, average male person in the UK, is 176 centimetres. My heights measure between 150 and 2 meters 98% at the time. So most people sort of have a height around 176, that's where the most common height is. But it spreads and it's the same with every single parameter you can think of. You might have a statistic that represents it and for data where the spread is symmetrical about the central point, that's a mean, but for most things in the built environment they're best represented by a median or geometric mean, which, if the Never mind, I'm not going to go into that right now.

Ben: 13:05

But this spread in the possible values is something that is really important to understand because some of the time a value could be below average and some of the time it could be above average. And if we sized, for example, heating systems in a building using an average temperature outdoors, you're never going to have a heating system that's big enough to handle most of the cold days. So you tend to work with extreme values right at the very cold end of the spectrum, whereas when you size a chillar, you're working at the very warm end of the spectrum. So you need to understand that spread of information. And I think that most people will understand that in their professional lives and they'll certainly know that if I was to ask them whether they were short or tall, because they would know, relative to most people, whether their eyes are looking up or their eyes are looking down. So you may have an absolute value, which is represented by the mean and the spread of the mean, but your relative value then is how you compare against some threshold, and that's generally in height, what other people's heights are. So at the beginning of the pandemic the very long roundabout we're starting to ask your question.

Ben: 14:31

We started to look, we needed a model that would say how much virus is somebody emitting into the air? How much might somebody who was susceptible to infection be inhaling? Therefore, what was the probability that they then might become infected? And there is a very well-known equation called the Wells-Reilly equation, which was developed using measles and tuberculosis outbreaks, where much of the information that we needed just didn't exist for SARS-CoV-2, wasn't there. So we were having to artificially create numbers and it wasn't giving us anything that was particularly meaningful. So what we did was we were able to take a relative metric, which is where you divide. You take your numbers for a scenario and you divide it relative to a baseline scenario, and all the numbers that we didn't know dropped out of the equation they cancelled. So we're then left with an equation that's quite precise but inaccurate, but it's going to give you something that's more plausible than just guessing at numbers.

Simon: 15:55

And this is where you get down to the practicalities of uncertainty, because you said two things there that seem contradictory.

Ben: 16:06

If you imagine shooting bows and arrows at a target, if you get close to the bull's eye, then you're accurate. If your spread of arrows is right around that bullseye, then you're also precise. Now you could have a perfect spread of your arrows around the bullseye, but they might be not very close. They might be, you know, far out. So then you're precise but inaccurate. Sorry, wrong way around. You're imprecise but accurate.

Ben: 16:40

So what we've done is the precision then is an indicator of the spread of the data. So you're imprecise as a huge spread in your data. So what we're doing is reducing that spread. But we still don't know how close to the bullseye we are. So what we were able to do is to say well, if you increase your ventilation rate by a number of folds, your percentage reduction in infection risk would decrease by a number of percent. But if you've got very little virus in the space to begin with, a reduction is, you know, a large reduction of a small number is a small number. Similarly, a reduction of a large number may not make much of a difference either. So you never quite have a handle on just what you're doing, but you can have a percentage reduction.

Simon: 17:32

And ultimately, I guess for the layman, the hunger for certainty is unrealistic, because in a real environment you can apply a good formula to something, but if you don't know how many infectious people are in that room at that precise moment, how many people are vulnerable at that precise moment, where they're located at that precise moment, there's all these uncertainties baked into the real world. So it's quite hard for laypeople like ourselves to understand how you translate formulas and maths like that into something that's useful in a practical way.

Ben: 18:15

So to us. So a model should only be as complicated as it needs to be, but because of the uncertainties, it meant that we automatically rejected the use of CFD, for example, which would give you a very discretized understanding of a space, but you don't know where the infected person is. As you said, if you don't know where your furniture is located, you don't know where the air is coming in or going out, and we try and generalize things. We ended up with a well-mixed model which is used to relate carbon dioxide concentrations to ventilation rates, for example. It's the same idea, but with some tweaks to account for the aerosols, and that means that we argued then that it could be generalized to all spaces, because the only mixing conditions that you could say occurred in any space is the well-mixed spaces. The only one condition that could be generalized to every space in existence, doesn't matter what it is. You can say that.

Ben: 19:25

And of course, there were lots of videos coming out about people coughing that were around at the time and people walking past each other and interactions. And of course, if somebody just gets up and walks across the other side of the room, then they throw up all sorts of turbulence around the room and it tends to the well-mixed solution anyway. So better to have a simple model that you have a handle on the uncertainties than to have a complicated one where you really didn't.

Simon: 19:54

Yeah, interesting. And ultimately that work led to you being involved in the development of the 2.4.1 standard, didn't it? Because you started to take some of those formulas and run them thousands upon thousands of times to come up with some of the standards we saw in 2.4.1.

Ben: 20:18

Yes, there was something you mentioned earlier too about about not knowing how many infected people were present in a space, and at that time we were just starting to think about that very heavily. And there have been community infection rates given by the government throughout the pandemic. So we knew that, based on those simple probability theory, we could start to work out whether people are present in spaces.

Ben: 20:44

Now, most of the models up to then had just assumed there was a single person infected in all spaces, which is always wrong. So for many of the conditions for the community infection rates in many of the spaces, the most likely number of infected people most of the time was zero. So it was only when the community infection rate went up, particularly above a couple of percent, that it became more likely there was one person in the space. And if you have a small space, that probability of infection is very low because there's fewer people in there. But if you've got a big space, like a theater or football stadium, then the probability of having infected people is almost certain.

Ben: 21:32

So why did we account for that? Well, we started to account for that in our models and I think it was some of these discussions then that we'd started to publish and, throughout the interest of ASRAE, had to apply. You know, we weren't chosen, we weren't. We didn't get a tap on the shoulder or anything. We were asked to apply and I did, and there were hundreds of applicants and we were selected to participate in the modelling work package and I was actually vice chair to Marwa Zatari of the modelling working group. So all those models then were used in various forms to help underpin the ECA rate, quivalent clean-in rates. That's right yeah.

Simon: 22:22

Which is important for people to understand. Generally, when we're talking about ventilation rates, we talk about a clean air rate, and that that's usually has come from the amount of outside air we bring into the built environment. So what is the fresh air delivery rate, the air change rate, into that space? But what 241 did, which was key, was to understand that it may not always be possible to bring that much fresh outside air in or outside air in for a whole host of reasons. But there are other ways of achieving an equivalent to an air change rate and that that's what that describes.

Simon: 23:01

Is that maybe a mix of mechanical ventilation with some fresh air delivery but some recirculated air with some form of cleaning, whether it's filters or UV, for example? So I'm guessing your work in 241 was trying to develop what those kind of levels would look like, and I remember saying to Max a couple of weeks ago that some of those levels seem quite eye-watering to us over in Europe. When you start seeing 20 litres a second plus per person, they almost seem unattainable. Actually, when we look at the existing state of our built environment that the prospects of delivering 20 litres a second per person seems extraordinary in a classroom, for example, you know where we wouldn't get anywhere near that typically.

Ben: 23:52

But a good school classroom shouldn't be sized to provide the air quality flow rate. It should be sized to provide a flow rate designed to dissipate heat gains and mitigate against exposure, a sort of exposure against overheating in the summertime. So many of the classrooms in the UK that are naturally ventilated should have that capacity built into them. But of course, one thing that the standard allowed was you to drop your occupancy. So if you really couldn't meet it and you couldn't get hold of air cleaning technologies to add to your ECA, your air cleaner rate, then you could just reduce the number of people in the space and that of course has the effect of reducing the probability of the presence of infected people.

Simon: 24:43

Yes, yeah, indeed. So there are those kind of formulas baked into 241 where you can play around with and say, OK, if I can't quite get here from an equivalent air change rate, I might just reduce the number of people so that I'm able to achieve the it allows you to can do. Yeah, yeah, that's the interest. I think that was what Max was trying to get across in our discussion. Was this? It's not 241's job to tell you what to do.

Ben: 25:11

Yes, If we go back to that though, simon, back to our simple models we were able to use by the time we got to the end of, by the time we got to the beginning of looking at 241, we started to have an awful lot more information about the virus in its various forms, so we were able to get a better handle on the probability of infection. So we were starting to use, too, some clinical data, which was particularly interesting, and what we were able to find is that the uncertainty in the virus emission rate is all does magnitude in uncertainty, and that's a. We found it was at least sort of five or six, and that's like going out to measure a millimeter and coming back with a kilometer. So when you talked earlier about doing tens of thousands of calculations, yes, that's exactly what we did in order to try and work out what the effects of that uncertainty are, and for every single scenario we did 10,000 calculations, 10,000, so a lot. So we had a jolly good understanding of what might be going on, we think.

Simon: 26:31

So back, I suppose, to my original question, which was how do we apply those lessons on managing uncertainty into the real world?

Ben: 26:44

There's a really good Wikipedia page. I don't often recommend Wikipedia pages, but there's one on orders of approximation and we're really operating at sort of first or second best order of approximation most of the time. At first order of approximation, which is pretty rough, that's one significant figure or two significant figures and if we start to think like that we may be a little more skeptical of the numbers that we are producing. I don't think we can have too much faith in numbers that we produce. Really, everything should be given to no more than two significant figures in many circumstances in the built environment.

Simon: 27:35

But it's what we have to work with. That's the reality.

Ben: 27:38

Right indeed, but we have to make decisions with honesty, without saying I totally believe this value, because I'm not sure I always do. I think it's like I said the probability of there was 100% probability that somebody was going to be infected was between 0% and 100% in any space, and sometimes that's the best you can do.

Simon: 28:05

Yeah, indeed, and also we work in an environment where people are in control of their own environments and their tolerance of risk may differ as well, so we can struggle with uncertainty at the calculation level or at the monitoring level or the kind of numbers that we can present to people. But we're also dealing with behaviors and habits and people's tolerance of risk, and how we frame risk in the built environment is a really key part of that. So much of that work was done during smoking and anti-smoking. I always say to people that everybody knows what a dali is. They just never heard it called a dali before.

Simon: 28:50

But we all know that every cigarette costs you seven seconds of your life or something like that. We've been using that kind of language in public health for a very long time. Both your work on COVID and on these harm intensities ultimately has got to translate to behavior change and changes in regulation and things. Do you see this kind of work sitting in that realm of anti-smoking and framing of these problems in a completely different way so that people can get their head around them? Because we've got to start viewing the quality of the air that we breathe in the same way that we view the water that we drink and the food that we consume. We just haven't been very good today to communicating that well. So some of this work is the start of that process.

Ben: 29:45

I hope so, because I really think this is a possibility to have a really great effect on a huge number of people for very well, possibly some very small changes. Let's take, for example, the outcomes of the Nottingham study and say we accept that particular matter is the most harmful contaminant by an order of magnitude. How can we mitigate against that Using a very small change? Well, building regulations changed a year or so ago and they were great and they allowed us to. They insisted on the cooker hoods in every new home and that they made it a very clear distinction between cooker hoods that recirculated and cooker hoods that extracted to outside, which is really important. We want one that extracts to outside.

Ben: 30:38

Another small change would be, wherever possible, whenever you have a new kitchen, what about we get a cooker hood in? If you can't get a cooker hood in, what about you get a hole in the wall fan? You know, just a basic extractor fan that is there, and that isn't something that government has to pay for. That's something that the consumer pays for. It promotes technology. It promotes us to spend a bit more money on a cooker hood in our homes and hopefully use it. If you want people to use it, then we need innovation to make the things quieter.

Ben: 31:14

I like cooking very much, but I also like listening to the football while I cook, and I remember, before I knew all this stuff, I used to turn the cooker hood off because I couldn't hear the radio. Now I listen with headphones, and noise-canceling headphones at that, which is a bit of a problem because I can't hear the food, but at least I know that we're a little bit safer. So point is you can have small changes in a lot of homes that affect a lot of people and you will see, in all probability, positive health effects in decades to come.

Simon: 31:53

And opportunities for innovation in the supply chains and marketplaces. I mean, I can't recall, outside of, perhaps, commercial cookahoods, the specifications and performance of a cooker hood being advertised in any way or being sold. Even, like you say, I mean, most cookahoods are absolutely dreadful from an acoustics perspective and a performance perspective. So this huge opportunity for industry to innovate and start producing cookahoods that are quiet and deliver good capture efficiency and turn on automatically when you start cooking and doing all of these things, that would make the impact on health at a population level Very, very impactful Possible.

Simon: 32:41

Yeah, because at the moment, like you say, most people, apart from turning it on for a steak or something big in our kitchen, the cooker hood doesn't come on. It sounds like a jet engine, you know. So, yeah, there's an almost room for innovation, I mean on that kind of future tense. To finish off, ben, if you're kind of advising the next set of researchers and academics coming through, where would you be pointing them to for the next tranche of research and study? I know it's a very difficult one to answer, but I'd be interested to see where you think the areas of innovation in air quality and ventilation are going to come in the next few years, I think we've touched on quite a few.

Ben: 33:29

So we've touched on technologies needed to mitigate against exposure. For example, some building types are already mechanically ventilated. Is it possible to put a filter on the flow coming in so that we filter out what's outdoors before it comes indoors? Particularly the particular matter that's relatively straightforward to do. What do we do in naturally ventilated buildings, particularly those that are in industrial areas near big road dunctions, where there are vulnerable people and maybe children or elderly or whatever? What do we do with those? So then there's an air cleaning approach that could be thought through to mitigate against that. Or is there the retrofitting of mechanical systems and maybe a wall-mounted box that provides what might ordinarily have been put in the roof or stunts centrally?

Ben: 34:26

I'm a big picture guy, so I'll be looking at big picture stuff. I prefer to do that. So I'll be trying to understand the bigger picture for public health and how the built environment fits into that. And I think that we tend to think of the built environment in a very isolated space and I would like to work out how it fits in with, say, transport or other environments, so that we can understand our place in the world. And then there's questions about what do we do for the next pandemic? Are we ready for that? Is the information that we've acquired for this pandemic suitable for other airborne aerosols? Maybe, maybe not, and I think that needs to be thought through. We've talked a lot about physical health. We haven't even mentioned mental health yet, and I've not realized that's a gray area, but I think that's a huge area for knowledge development over the next few years. What does an engineer need to know about mental health and the built environment, the relationship between the systems, the services that we provide in buildings and mental health? I think that's a big knowledge gap.

Simon: 35:45

Really good point, I think it's. Joseph Allen says that your building manager probably has more of an impact on your health than your GP over your lifetime. I think it's an interesting way to view this that the general health and well-being of people in the built environment is going to have to change and how we view it. So I think this also brings in social sciences and behavior and communication. There's a lot of work to do, so I think it's going to be an interesting next decade or so.

Simon: 36:19

Ben, listen, thanks so much for your time today. It's been a pleasure to talk to you. As always, it's been really interesting. I think the work on harm is fascinating. It's going to be great to see what happens with that. So, ben, I appreciate your time. Thanks very much.

Ben: 36:35

Thank you for having me.