The Demonstration of Energy Efficiency Potential Project Findings

00:00:00:00 - 00:00:23:03

Speaker 1

Good afternoon, everyone. Thanks so much for taking the time out of your day to join us to talk about this fascinating project. I am Morwenna Slade, I'm head of historic building climate change adaptation here at Historic England. I'm based in the Technical Conservation Department, and I lead a multidisciplinary team focused on the adaptation within the built environment.

00:00:23:13 - 00:00:51:04

Speaker 1

We undertake research, write technical guidance and support a range of projects and workstreams in all different varieties. And one of the many things that my team gets up to is being involved with projects such as this either to a great extent or a small extent. It's a real pleasure to be able to support the dissemination of the findings of the DEEP study.

00:00:51:21 - 00:01:18:02

Speaker 1

Joining me today is Roger Littlewood. He's from the Department of Energy Security and Net Zero. Roger is based at the Energy Research Team and provides technical advice for policy teams, as well as leading a range of research initiatives, including this project. Also joining us today is Professor David Glew, the director of the Leeds Sustainability Institute and head of Energy Efficiency Policy.

00:01:18:20 - 00:01:57:05

Speaker 1

David manages a range of large interdisciplinary retrofit evaluation projects and undertakes building performance evaluation. He also looks at occupant comfort and behaviour change, which is, of course the the important other side to a lot of the modelling or building performance evaluation. As Roger and David will discuss, this is one of the largest research projects into solid wall retrofit undertaken and it has successfully demonstrated that the whole House approach is a fundamental part of reducing risk, highlighting that solid wall insulation is an important part of the solution.

00:01:57:14 - 00:02:21:14

Speaker 1

It directly aligns with much of the work my team does and will be doing over the next few years to provide the best practice, advice and support decision making to reduce and avoid risks once we insulate our homes and really try and support people to make good decisions within the built environment. So now it's my great pleasure to pass over to Roger, who will introduce the project, and David, who will tell you more about it.

00:02:21:24 - 00:02:24:21

Speaker 1

Thank you very much.

00:02:26:10 - 00:02:57:09

Speaker 2

Thank you very much, Matt and Morwenna for this opportunity to speak and for the intro. So yes, as Morwenna said, I’m Roger Littlewood from the Department for Energy Security and Net zero, managed the DEEP project since it started and as Morwenna said, I’m in our energy research team. A very brief intro before Dave talks about the detail of the project, but a bit about the project's whys and wherefores from me first. DEEP is a multi-year

00:02:57:12 - 00:03:33:06

Speaker 2

It was all intended multi-year, but it's being longer than longer than we originally intended. But it's a research project into domestic energy efficiency, wholly commissioned and funded by DESNZ energy research team. As I said, where I sit, the. As to why it commissioned the research in the first place. Well, prior research in, into domestic energy efficiency in the UK is mostly studied retrofit measures in isolation, leaving gaps in understanding and to the potential of retrofit being unfulfilled.

00:03:34:01 - 00:04:13:07

Speaker 2

What we wanted to do and Dave the team have since taken this on delivered is take a system level holistic approach looking considering both the fabric building, fabric ventilation occupants and other factors as well frankly as part of the whole house approach to retrofit which against primary risk reduction and as Morwenna said and the opportunities and risks are of of not taking that so perhaps taking it versus the piecemeal approach, i.e. installing measures one by one without due consideration for how they can interact and overlap with one another.

00:04:13:14 - 00:04:39:08

Speaker 2

And like any good research projects, and we want this to provide an evidence source for future policy standards, paths and indeed future research and innovation work in the field. We commissioned a project and awarded the contract to a consortium led by Leeds Beckett University, supported by the Universities of Salford and Loughborough. They will talk about more and more about that in a moment.

00:04:40:02 - 00:05:06:24

Speaker 2

And the they've also been supported by subcontractors and landlords and other stakeholders. Besides, there's been a long running I would say it's been a long running projects having begun in 2019, originally due to complete in 2021, but extended on a couple of occasions, partly partly down to product specific circumstances and partly due to more global things like COVID lockdowns and supply chains getting hold of materials and things like that.

00:05:06:24 - 00:05:29:00

Speaker 2

But we very much hope to publish the reports in the next two or three months or certainly before the end of the year anyway. Now, right. So as to to extend the for the how of the project, what the team has done and what they found or is least a flavour of. That's to hand over to Professor Dave Glew from at Leeds Beckett University.

00:05:29:00 - 00:05:55:06

Speaker 3

Thank you. Super! Thank you very much. Right. We’ll just crack on with it really. It's been introduced very nicely there. So this is Leeds Sustainability Institute just before the obligatory this is who we are in case you don't know, this is just taken from our website and you can just Google us if you don't know who we are. We do research in sustainable buildings, urban environments and behaviour, and we've got a legacy of 30 odd years of research.

00:05:55:06 - 00:06:24:10

Speaker 3

But if you're interested, you can find out more. But we're here today. I have the real pleasure to talk about the project. I hope it's going to be a pleasure to you all listening as well. What did we do? This is going to talk about why we did it, what we did in terms of which homes we looked at, what retrofits we did, and I'll move onto the results was effective in terms of the different reductions in heat lost that we found in these solid walled homes or not.

00:06:25:02 - 00:06:42:24

Speaker 3

Then we're going to move on to okay, that was what we measured. But what about the models? How did the measurements compare to the models? And that's got implications for things like EPCs, which I'm sure you are all in love with as much as everyone else. And then we going to very importantly talk about the risks of retrofit.

00:06:42:24 - 00:06:59:01

Speaker 3

And this is one of the most important things with with the whole House approach, and that's why we've done the project as well. If there's time, this is a monumental set of slide back here. We will do our best to get through to some of the interesting stuff that we found, which I hope you might find interesting too.

00:07:00:13 - 00:07:26:22

Speaker 3

This is the consortium it's been described. We're really grateful to the Department for Energy for funding this over four years. And to our partners, especially landlords, who at a time when people need homes and houses, they were very grateful. We're very grateful to receive this support, to go and actually rent out these homes and insulate them and measure the performance of homes from all these different landlords, which is great, allowed the project to take place.

00:07:28:02 - 00:08:03:21

Speaker 3

Okay, so Leeds Beckett, we went out and did a series of retrofits across 14 homes and what we then got and we would measure and monitor and model the performance of these homes and the retrofits for that performance and risk. Salford then replicated one of these case study retrofits in their energy house to see if we've got the same sorts of results in a laboratory environment as to those that we were experiencing in the field, because you get a lot of noise in the field when you're doing building performance evaluation, I'm sure lots of you know, and then Salford, Loughborough and ourselves, we did lots of energy performance.

00:08:03:21 - 00:08:29:06

Speaker 3

Did the models match what the measurement said and what where our models risks and how did they match what we were measuring in the field to help them out? We got some brick samples from some of our homes to calibrate our models to make them a little bit more accurate. I could talk about high level. And then aside from the case studies, we also were interested for the project to look at what sense of potentially the energy performance of homes in general is.

00:08:29:06 - 00:08:48:00

Speaker 3

What can you pick up just by doing some surveys and specifically, what's that tightness of homes in general across the UK? So we did lots of tiny surveys and lots of types of homes and I'll show you a brief set of results about that too. Okay. That's the totality of the project in terms of what it was, all the activities that happened.

00:08:48:11 - 00:09:08:06

Speaker 3

And as I mentioned, the whole House approach was fundamental to the design. But what is the full house approach? So this is just a schematic diagram to show that the whole house approach doesn't mean insulating the whole house. So we have to get that out of the way straightaway. Be very clear about how you can install a single measure via a whole house approach.

00:09:08:21 - 00:09:31:20

Speaker 3

And all that means as you consider how the thing that you're doing affects the rest of the house and all of the risks in the home. So we've observed previously that lots of retrofits are happening in piecemeal, so they're not considering how they interact with other insulation or interact with the ventilation in the property. And those are perceived to have manifested high risks.

00:09:32:05 - 00:09:49:03

Speaker 3

So we want to see people doing whole house approaches where there are lower risks. So we're agnostic to the actual measures that are taking place. Whatever's happening in the home, we want to make sure that they're in a whole house approach. They bring in those people in the top two boxes down to the bottom two boxes. Okay. So that was that.

00:09:49:03 - 00:10:13:13

Speaker 3

That's the concept of the whole House approach for those of whom we find there's a lot of confusion about people, assuming it means doing every single thing you can possibly do, it really doesn't mean that's all. Okay, so that's the concept. A, what did we actually do to investigate that? Well, we would take a home which usually had loft insulation already indicate by this lovely diagram we would then deliberately install a retrofit measure in piecemeal.

00:10:14:02 - 00:10:31:10

Speaker 3

So we would deliberately do it the way that we are encouraging people not to do it. Okay, so let's take this instance. We did an external installation. The reason we did it like that is it would give us the perfect opportunity to measure and monitor the under-performance and the additional risks that occur when you do it like this.

00:10:31:23 - 00:10:56:09

Speaker 3

And then, of course, we could go back and convert it to a whole house approach. So in this instance, that would mean, for instance, ensuring a continuous insulation layer around the whole fabric. Great. So now we've got those scenarios where we've got the base case, the piecemeal approach and the whole House approach that we can compare across the different performance and risk profiles of those things.

00:10:56:09 - 00:11:19:23

Speaker 3

So that with the projects these are the homes. So we've got mostly brick homes lost, a nice age range of homes, a couple of non trads, stone property in here too. So out of the 14 we've got a really nice mix of ages and house types and we've also got a good array of different types of retrofits in the homes too.

00:11:20:04 - 00:11:47:24

Speaker 3

So you can see that we've installed insulation in five of those homes. Just proofing has just been cropped, but it's nine homes and then loft for greenery for for lots of new glazing. And that also includes new, new external doors, which is important. Ground floor we did twelve, but two of those were solid and ten were suspended. And then that whole house approach that deliberately did that piecemeal and then come back and do those tidying up jobs to make sure it's a whole house approach.

00:11:47:24 - 00:12:05:00

Speaker 3

Did four of those okay. So a good range of homes, a good range of retrofits. So that would look something like this. We would go and do a test on our home as we found it that would have a little bit of loft insulation. Typically we would then either upgrade or replace the loft insulation, do all our tests and models again.

00:12:05:00 - 00:12:25:23

Speaker 3

So we've got our first comparison in our step. The third step would be to do the glazing again. We do all of our tests and all of our models. We then do the floor, all of our tests and all of our models, and then the Swg in this instance, it was external, all of our tests and all of our models and then in on the red around that which represents the external slash, you can see isn't pink.

00:12:26:10 - 00:12:49:10

Speaker 3

And those represent those whole house approach details that might conventionally be missed from time to time. So below that of course is commonly excluded from from schemes and installing external insulation also at the eaves, whether it's calling architectural details, often I stop short there or we have a detail like a bay roof and what it means as an obstruction to installing insulation.

00:12:50:01 - 00:13:07:12

Speaker 3

These are the sorts of whole house things that we were interested in looking at. Okay, so that's typically what we would do. And here's just a deep dove into one of those just to show really the fundamental task behind what we were doing. So for energy performance, it's something called co-heating test. Some of you may have heard of it.

00:13:07:12 - 00:13:35:09

Speaker 3

It’s a way of essentially measuring the aggregate energy consumption or and efficiency, I should say, of a home and the units, it has a heat transfer coefficient. Okay. So the high or HTC the worse your home. So we want to be seeing each retrofit stage reducing the HTC that we were measuring. So we started it with this house around 250 watts per Kelvin and then we'll have it when we show the loft insulation very little really not a huge impact but that's that's fine.

00:13:35:09 - 00:13:57:11

Speaker 3

Already had some loft insulation we might have predicted that we then put the glazing in the new external doors on and again another small drop and again that's a small surface area of the whole house. So we're measuring a small reduction. But what's important is it's out, it's within the test uncertainty measurement. So we can't yet be sure that we're actually achieving measurable savings.

00:13:57:11 - 00:14:38:09

Speaker 3

We're achieving savings. We've improving the u-values. But actually we're not not be able to measure the statistical significant difference between the different retrofits. Same scenario for the floor. It's a small area hasn't reduced a significant change. It looks like it goes up, but that's just indicative and it is actually no change is is that because it's within the test uncertainty and really this is just to highlight that it's only until you insulate the solid walls that you get a fundamentally large drop in the energy and the heat loss from the homes is super important to address.

00:14:38:09 - 00:15:01:04

Speaker 3

The heat loss through the walls in these homes. If we're serious about helping the people inside these homes, okay, that was one case. So let's have a quick look at all of the homes. Put together. This is what it did in terms of the reduction achieved in the HTC measurements. Again, so this is a graph which shows all of our different retrofits across all of the different homes.

00:15:01:04 - 00:15:23:06

Speaker 3

So each blue bar Hastie's heat transfer coefficient just seen on the the chart, it's how much energy it costs to heat the building by one degree. That makes sense. If he's interested, you can look it up. What we can see here is all different blue bars. Now, the size of the blue bar indicates the uncertainty associated with the measurement.

00:15:23:24 - 00:15:46:15

Speaker 3

Okay, so the bigger the blue bar, the more uncertain the test, the amount of uncertainty in the result. If the blue bar is touching the zero line, it means there's no statistically significant result. So although it looks like there's lots of blue lines below the zero line, actually they're all touching the zero lines. That just means we can't tell with the test method if there is a change or not.

00:15:46:15 - 00:16:11:13

Speaker 3

But there is. There are some bar blue lines that are significantly above the zero line. Solid wall insulation every single time is making a statistically significant difference. We also get a couple of others say one floor suspended floor, one ground floor, solid floor, one loft and one draft proofing measure. Also had significant improvement that could be measured, significant reduction in HTC.

00:16:11:19 - 00:16:36:08

Speaker 3

All of this may have achieved a reduction but this is what can be actually measured. Now this really just shows two things real difficulty measuring performance changes in single homes with even the most advanced scientific methods available, but also the fact that the variation in performance of retrofits is very large. And that really depends on the base case of your home as well as what retrofit you're installing and specification for the retrospective installing in these homes.

00:16:37:03 - 00:16:57:03

Speaker 3

Okay. Okay. Well, that's the measurements for what about predicting the model. So if you can't measure it with a 100% certainty, what can we measure? So this is a graphic I'm going to ask you. You could probably ignore the blue and the orange lines at the moment. They're just different types of models. And come on, today's in the minute, the green line is your EPC prediction.

00:16:57:12 - 00:17:19:12

Speaker 3

So what we'll see here is the models are pretty much confirming what we're measuring, i.e. all of the other types of retrofit, the draft proofing, the loft, glazing the floors. They're not really making much of an impact on the energy efficiency of the home in terms of substantial fuel bill savings. It's only when you insulate the solid walls that you get serious amounts of savings.

00:17:19:12 - 00:17:36:11

Speaker 3

It's being achieved. A couple of quirks in here, the zero hour tightness benefit because in RdSAP and EPCs, you can't assess the air tightness of a home improve the airtightness of a home. It’s just a quirk of the model and also you'll see the room and roof has got really nice strong green bars. Now that's not because that's being achieved in practice.

00:17:36:11 - 00:17:54:20

Speaker 3

It's because of a very poor set of assumptions that back up the pieces assumptions on heat loss from room and roof so those that we're going to talk about those in a minute okay so the models are sort of confirming what we're finding that really is solid, also the most important, but also that there's a huge amounts of variation in performance across these different retrofits.

00:17:55:06 - 00:18:15:11

Speaker 3

So why is a such a lot of variation? Here's a really nice scenario we threw at six of our homes. We're actually passed three paths, so these are identical, so inverted commas homes that we would really expect. And certainly the model would predict these times to have an identical benefit in terms of the retrofit that we're having. What we find, of course, in practice is identical.

00:18:15:11 - 00:18:45:06

Speaker 3

Homes are not identical at all. These variations in terms of layout floor areas in these homes which are causing and driving quite significant changes in the predicted performance and the measured performance and also homes have been messed around with in the past. So this is just some images to show that we found multiple instances where landlords didn't know that some of their homes had some walls which had polystyrene, plasterboard, insulation in them, and they didn't have a clue what was behind their sloping ceilings.

00:18:45:06 - 00:19:09:21

Speaker 3

For instance, in the homes. And the thermal images on the left are in the same house. Only one of the rooms had had some kind of IWI and the other room hadn't. So with all of this variation, how on earth are you supposed to predict what benefits you're going to get and how on earth you supposed to design the appropriate retrofit solutions when it's really difficult to even understand what's going on in the homes in the first place.

00:19:09:21 - 00:19:32:10

Speaker 3

So these are some real barriers and one of the reasons why it's really tricky to understand exactly how much benefit you're going to get from a retrofit before you start. Okay. Well, we did have a go. We know we're using some pretty advanced intensive research methods here. And what we've got is and what we found is that even when you've measured a home, you've got to be incredibly careful.

00:19:32:10 - 00:19:56:15

Speaker 3

So we've got we plastered our homes full of people experiences. So this this is a device which will allow you to derive the u-value if you element's. And we were able to then improve our predictions on how the performance of the different retrofits by understanding the background thermal performance of our base case homes. But what we would find is that even on the same element, you would get a different u-value depending on where you placed your sensor.

00:19:57:02 - 00:20:26:19

Speaker 3

So, you know, word of caution, yes, it's great that we can start introducing new measured values into our models, but you might end up in a scenario where you're putting in the wrong number into your into your model just because you've placed a sensor in a particular spot. So we overcame that by placing multiple sensors, but that's incredibly resource intensive thing to do and perhaps not practical, but something we did in this project, I know sometimes we would find that the same element might have 100% different u-value.

00:20:27:04 - 00:20:51:12

Speaker 3

But depending on where we place the sensor, for instance, okay, so this is a graphic now which shows you what we did measure and what was modelled. So the important thing to to show here is there's then there's some missing values, which is a shame. So there's the green boxes are what the model says. The blue boxes are what you get if you use the battery calculator.

00:20:51:14 - 00:21:08:22

Speaker 3

So you kind of knock a hole in the ceiling and see what's there and you create your own calculations evenly. And then the orange boxes are the the things that you measured. So in the last slide, we use those four places, those does the variation in the performance is where you value the walls that we measured in the homes.

00:21:08:22 - 00:21:33:02

Speaker 3

So that was 31 different wall types across the 14 different case study buildings. So that's why this that we want external measurements and what you can't see there is the u-value of the green, which is the EPCs prediction. Now, most of you probably know that the EPC, EPC assumes a solid wall has a u-value of 1.7. Okay, so just imagine a line at 1.7 for green.

00:21:33:21 - 00:21:52:10

Speaker 3

And one of the biggest findings really is that when you measure u-values of solid walls out in the field, there's a huge variation. Yeah. So we were measuring some some at u-values above two and somehow devalues below one. Again, that's going to have a really profound effect on your models and on the performance of the retrofit that you're installing.

00:21:53:10 - 00:22:13:17

Speaker 3

Same thing for the sloping ceiling. And just imagine that there's a a line at 2.3. So the EPC assumes that all sloping ceilings have a U value of 2.3. In reality they don't. So these are some of the variations in terms of kind of like assumptions being take taking place in EPCs that we can we can improve on. So that was a useful finding.

00:22:14:10 - 00:22:44:04

Speaker 3

Okay. So now we insulate these walls and floors and ceilings and hey, presto, they all drop. So great news you our insulation is achieving tremendous u-value savings for each of these different elements. Perhaps more importantly is we now have alignment pretty much with our models and our measured values. So post retrofit, we're not too concerned about the model inputs being used in our models because they're roughly aligning with what we're measuring.

00:22:44:11 - 00:23:08:09

Speaker 3

And that could be predicted because of thermal resistance provided by insulation is obviously dominating the overall U-value. Okay, so that's that was good news. That's heat flow through the fabric. And now what we've got is heat flow from background ventilation. They said the cracks in the gaps in people's homes. So I'm sure you're all aware with the blow a door test, which is what we used to measure the air tightness in these homes.

00:23:08:09 - 00:23:30:21

Speaker 3

And this graphic shows what we did in a series of homes to try to reduce the amount of air leakage. Okay. So there's 6 homes 7 homes here and it shows the journey of the air tightness as we did different things on those homes and the story here really is that sometimes we got a tremendous benefit by just doing some simple draft proofing.

00:23:31:18 - 00:23:57:12

Speaker 3

General sealing mastic comes and removing redundant ventilation that was installed for things like gas fires that are no longer there or just boxing in new pipes. So that first home we went from 50 metres cubed per hour down to ten, you know, great saving. But then the next two homes, which were at the same starting air tightness as the first home, nothing, no change whatsoever.

00:23:58:04 - 00:24:31:17

Speaker 3

So we didn't investigate the more advanced air tightness procedures, you know, stripping back walls, applying membranes across surfaces, taping up between different elements because we were interested in what currently happens. If you ask a draft proofing specialist to come around a decent work. So I guess unfortunately one of the findings is you don't know in advance if it's going to be worth it and if you're going to get any savings in terms of air tightness, when you when you use these these sort of very basic tightness or draft proofing, we should call the measures.

00:24:32:04 - 00:24:53:02

Speaker 3

The final finding here is that carpets made a tremendous benefit to the air tightness of the home. So if we're measuring the air tightness of the homes before we finished fitting them, before we finish the floor coverings, we're probably getting a false value of air tightness. So we need to be careful about what we use in that value for. Okay, so that's draft proofing which had variable success.

00:24:53:02 - 00:25:17:13

Speaker 3

So as we're talking about air tightness at this point, I'll just kind of go to those hundred and 50 homes that I mentioned earlier, which is kind of separate to our retrofits. And this is just the profile of the air tightness that we measured. So the red line is building regs eight. And as you can see, most of the homes are above that fine line, which I think we would have expected. SAP actually you could draw the line at 15.

00:25:17:13 - 00:25:47:16

Speaker 3

SAP tended to assume most of our homes had an at times of around 15. And clearly you can see most of our homes don't have that tightness of 15 is much lower than that. So potentially we are over predicting the amount of air leakage in homes in EPCs, but then probably the work near, the saddest stories, those homes at the right hand side, whether they've got a real problem with excessive air leakage and there's really no point insulating homes that have access to that leakage unless you're going to sort out the air leakage as well.

00:25:48:00 - 00:26:13:20

Speaker 3

And so there's no mechanism or policy for these people to kind of going back behind this, nothing for them, because EPCs don't consider that tightness in improvements or bespoke values in their calculations. So that's that's a real concern. I personally can be done about those those people. Okay. So there's lots of retrofits that we did and the benefits we also measured the costs, a word of caution before you take too much away from this slide, we were one offs.

00:26:13:20 - 00:26:32:13

Speaker 3

We were a university, one off kind of set of retrofit. So I don't think we can say our costs are representative of what you would get out in the real world when you're doing sort of neighbourhood neighbourhoods, street by street retrofits, economies of scale, etc.. So what we can so that is useful, I hope is the right shape between the additional work and the retrofit costs.

00:26:33:00 - 00:26:57:07

Speaker 3

So we were carefully kind of ensuring we kind of separated those two things out. And what we're finding is that at least a third of all the costs that are incurred were other stuff that wasn't really the fault of the retrofit. So for instance, in one of the homes where we were doing external insulation, we had to replace a garden wall that was owned by the Highways Local Council because it backed onto pavement and that cost £6,000.

00:26:57:13 - 00:27:22:11

Speaker 3

And so as soon as you start those sorts of numbers that that totally eradicates any sort of reasonable payback times that you could consider because of these extra costs that are taking place of the reasons for having real excessive extra costs were very common across the retrofits, was damp. So having to repair leaking roofs, missing bricks and then even things like clearing junk.

00:27:22:11 - 00:27:40:10

Speaker 3

So simple retrofits like loft retrofits could cut into maybe £1,000 or more because you need an extra escape, an extra couple of people for an extra day. And these sorts of things can really cause a problem to the payback to the hoping to get the retrofits themselves can actually cause damage. Say simple thing, is it simple to remove windows?

00:27:40:10 - 00:28:16:20

Speaker 3

And if the substrate starts cracking and crumbling, you end up deteriorating the opening and you've got to then make good and repair that. And again, that's kind of several thousand you weren't anticipating and not to say extra disruption and and hassle for the for the land for the homeowner. So it's all messy, really and quite and can be quite expensive and so take home so the actual stuff that we've done then I think the project providing a lot of evidence to kind of really if it's not already clever enough to really make the case that solid wall insulation is super important for tackling any loss in these homes, as much as good as all the

00:28:16:20 - 00:28:35:07

Speaker 3

other stuff can do, that's great. But if you want to achieve serious savings, you have to do the solid walls. I think that's a quite a strong message from the findings and having said that, the findings are incredibly variable, which is not necessarily good news because it means there's a lot of uncertainty involved in the benefits of these things.

00:28:35:20 - 00:29:00:24

Speaker 3

And so to try to minimize that, we think there needs to be much stronger focus on the surveys that we're doing before we then embark and design solutions and predict the savings that are going to happen and that's that's going to kind of help to reduce the uncertainty associated with these things. And are the findings really hard to actually measure the benefit of a retrofit in a single home unless you're using very expensive scientific tools?

00:29:00:24 - 00:29:22:17

Speaker 3

So that's got implications for trying to do mass use of smart meter data, for instance, to to to validate savings and that sort of thing is going to be very tricky to do well. And Yeah, in fact, we talked about the rest of it costs being quite large. Okay. So that's the test results we measured. How did the models shape up to those?

00:29:23:11 - 00:29:48:18

Speaker 3

So here's the graph for one home through its retrofit journey. Now the dark green line is heating value. That's our HTC again. And the green light green line is what the EPC is predicting. So straightaway, we can see we've got a real problem here at the EPCs massively over predicting the actual HTC being measured. Now that's also that's not only going to push this home into the wrong EPC band.

00:29:48:20 - 00:30:17:07

Speaker 3

So this is coming out as an E instead of a D, it's also making something called the pre bound effect. So it's going to think when you insulate the walls, you're going to get a massive saving, you know, hundreds of pounds per year. In reality, you might get 100 if you're lucky in this particular home. And so that's a problem for finance products that are based on retrofit savings over time, as well as expectations of consumers and the reputation of the retrofit industry

00:30:17:07 - 00:30:42:00

Speaker 3

I guess a large. Okay, so that was one home, but it was what can we do about it? So we've got a few things that we can do about that. So we've introduced two different coloured lines here, a blue line and then orange lines. The blue line is essentially the same as the green line. Instead of using the RdSAP software we broke in behind the scenes, we can kind of manipulate the calculation software.

00:30:42:00 - 00:31:12:03

Speaker 3

So although the blue line isn't perfect, the green line, it's pretty good. We got as close as we could get. There are a couple of assumptions around the way it calculates geometry that we couldn't replicate. It's being slightly different, but it means we can now make some alterations to the EPC model. The Orange Line is a totally different modelling software, so this is now using our dynamic simulation modelling, using our any time steps and it appears to be getting a better prediction of the measured value.

00:31:12:03 - 00:31:38:19

Speaker 3

So that's potentially one way that we could look at improving models is by changing the model itself, the way in which is calculated, that has some cost implications, but non-domestic buildings often use DSM for their building compliance model. So it's not a huge leap, but it would have some practical implications. And so I think there's a couple of bits missing.

00:31:38:22 - 00:32:08:09

Speaker 3

Okay, it doesn't matter. We can move on. And so what we can do is another way of improving predictions in EPCs is by incorporating the measured data that we that we have from from all tests. So the green line at the moment is all those green dots. That's our EPC values for all of our homes. Okay. So these are every time we did a test, we plotted the modelled predicted EPC, HTC against Arco heating, HTC.

00:32:09:05 - 00:32:27:18

Speaker 3

So really what we want is all those green lines to be populated along that black dotted line, the 1 to 1 relationship. So we wanted all those green lines to be on the black dotted line. That would mean our models were perfectly matching our measurements. Okay. And they're not. They're over predicting. So that's kind of what we've already explored.

00:32:27:18 - 00:33:05:11

Speaker 3

And then if we move to adding in the air tightness, adding in the u-values and adding in thermal bridging, we can make that prediction better. We bring that line closer to the black dotted line. So we're getting a bit better in terms of predicting our EPC can be improved somewhat, but it's still quite a long way. We're still over predicting our EPC quite a lot using the completely different model with all of those inputs as well shows that the design could actually predict those and measured values with a little bit more certainty.

00:33:05:11 - 00:33:24:03

Speaker 3

Still not perfect, too low variation around the line, but it seems that we could improve our models by improving our inputs, by improving the model itself. Okay, this is just a table to summarize that. So and it seems a bit complicated, but there's three different models. That's the green, the blue and the orange model in on the outside is green.

00:33:24:03 - 00:33:47:08

Speaker 3

Sorry. And what that first line is saying is that when you just do your basic EPC using all of the default inputs that the EPC assessor would just do if they just turned up on your curbside. We think on average EPCs for our case that he was over predicting the HTC by about 40%. That's really big. So we are really over predicting energy consumption.

00:33:48:02 - 00:34:11:05

Speaker 3

We can then by introducing all of these different known values which can be quite complicated, we can explain that a little bit later if you want to. We can reduce that over prediction level by introducing measured data down to about. If we look at the breakdown column, we can go from up 34% over prediction by using all of the defaults down to roughly half that, say 17% over prediction if we're using all of the defaults.

00:34:11:24 - 00:34:37:09

Speaker 3

Okay. So that's that's an indication of what we could do. The problem with that, though, is it those numbers in the brackets? So that's the range of performance. So it might not be a useful number to look at the average age to see a prediction, because in some homes, if we go back to the EPC, although the average is 40% of the prediction, some of the homes were over hundred percent over predicted and some of them were predicting less than the measured value.

00:34:38:06 - 00:35:07:08

Speaker 3

So that uncertainty, the variation in performance, it it's just people who have done a lot of research in homes know how quirky and unique every single home is. And it's just a representation of that buying. By looking at models and how things can vary dramatically when you incorporate different elements. Okay, so the modelling, what have we learned from the modelling while we're systematically overestimating EPCs.

00:35:07:09 - 00:35:30:11

Speaker 3

I think that’s quite clear now and there’s some corroborating evidence coming out whether institutes as well looking at that sort of stuff. Problem is it also means we're overestimating our retrofit savings. It's the rebound effect that we talked about and we can do something about it. We can add in real data, somewhat improves it, doesn't make it perfect. Or we could maybe look at change in the models that we're using by switching to DSM.

00:35:30:11 - 00:35:52:02

Speaker 3

But again, that's got its own implications associated with it. Okay. So that's that's the measurement to the homes. That's the modelling of the homes really exciting is to talk about the risks in the home. So we've got three types of risk settings condensation, institutionalization and overheating risk. Look out and start with surface condensation risk. So this is a graph of four of the homes.

00:35:52:16 - 00:36:14:15

Speaker 3

The blue, the green are in the elective names. And this is what happened when we started out with the pre retrofit. The bars represent the the the proportion of the home junctions junctions in the home, the risk of surface condensation. And we've assessed that according to having an fRsi below 0.75, which I'll just leave hanging unless you want to ask about.

00:36:14:15 - 00:36:38:03

Speaker 3

That's just the way in which we assess risk is there or not talk about it. Same. Essentially there's a huge amount of risk in the pre retrofit homes. The next set of bars, there's only three because we didn't do this retrofit in zero one by the next step this is when we did a piecemeal solid well retrofit said you remember at the beginning we kind of deliberately didn't include all of the junction details.

00:36:38:03 - 00:37:00:23

Speaker 3

This is what happens when you to the surface conversation risk in the home when you do a piecemeal retrofit so as a substantial improvement and that's a really important point uninsulated homes are unhealthy they solid walled homes are and how they are they have high risk of surface condensation and mould. You insulate them, it makes it better. But what you're doing is you reducing the overall risk for your local, the localizing it to your concentrating in a specific area.

00:37:00:23 - 00:37:16:22

Speaker 3

We'll talk about that an sec. But then when you are dropping, it's kind of below 20% of total junction lengths in the home. When you then get to the whole house approach, this is really good because you know, you're now below 10% at the junction lengths lower than five even in some instances, if you can get your detailing right.

00:37:16:22 - 00:37:41:03

Speaker 3

So this is really good evidence to show that, yes, the whole house approach is fundamentally a lower risk approach to retrofit. It's just one junction we're going to go into detail on and skip through the rest the eves. These are really common problems. So the base case scenario here, we've got surface temperature is really low. We've got below the temperature factor.

00:37:41:03 - 00:38:03:04

Speaker 3

So we're at risk of surface condensation. I mean, the whole wall is a risk of condensation here, but we're looking at the junction really common. We then see external insulation up to a point, but not then flowing the top of the junction to kind of match it with the loft insulation and all that's done. Yes, it's reduced. It's improved the scenario for most of the wall but we've concentrated some risk up at the junction.

00:38:03:04 - 00:38:24:21

Speaker 3

So that's a really common scenario. So we're able to quantify the risk that happens in that case. Then we would go back, remove the soffit detail on the outside, extend the insulation up through the back of that, and make sure that it's kind of more continuous insulation led to the loft insulation and what do we see? We've now removed that surface condensation risk by adopting the whole house approach detail.

00:38:25:05 - 00:38:47:21

Speaker 3

So in this instance, it's really nice evidence to say, yes, you do need to do this at this junction if you want to re remove the risk. I've said this condensation that but just being above a threshold .75 seems like a little bit of a simplified way of assessing risk. And there are the risks that we'll talk about in a second.

00:38:47:21 - 00:39:10:05

Speaker 3

But just looking at .75, this is a graph which shows the .75 red dotted red line and all of the different blue dots are our junctions. This is why they were assessed as sitting once we retrofitted them before we'd retrofitted them. So this is a uninsulated home and it's kind of scattered. There's absolutely tons of junctions here at risk below this .75.

00:39:10:05 - 00:39:38:20

Speaker 3

So let's see what happens when we do piecemeal retrofit. So this gets tidied up quite significantly that nice of the junctions and out above .75. But interestingly there's loads of junctions in between that range. What we've included there is kind of 0.05 range around this critical temperature factor because I don't think there's anyone here on the call would think that a junction that's got .75 and temperature factor is any less risk than a junction that's got .749.

00:39:39:09 - 00:40:00:00

Speaker 3

So, you know, I think having a hard threshold like that, I understand why it exists. You have to start somewhere, but we need to have a more considered on the standing of the fact that risk doesn't work like that. It's a continuous scale. So we've still got lots of of those junctions within that kind of potentially uncertain range.

00:40:01:08 - 00:40:20:01

Speaker 3

What happens then when we do the whole house? So we've gone back and we've assessed all of those different things. Most of these things are now above .8 even, and so we're really confident about those. But there are still a few that we're still within that range between .8 and .7 but, you know, that's a that's a much more successful outcome.

00:40:21:14 - 00:40:42:13

Speaker 3

Okay. We did that sort of analysis on the ease which you talked about in detail. Windows moving them into the EWI reveals when we're doing IWI, we looked at EWI cutouts, so things like gas pipes going up, walls, meters and stuff like that. The mounted onto walls and very common to see those not being removed by gas companies because that can cost thousands and take months to get organized.

00:40:43:11 - 00:41:06:02

Speaker 3

So we did some work on that. Looking at whether you don't do the deep sea insulation or you do do the deep sea insulation, something we looked at whether you have to insulate the suspended floor at the same time as insulating the walls or not. So we did different combinations and the benefit models is you can just run all the different scenarios that you can think of, the different combinations of junctions, which is what we did, which is great.

00:41:06:19 - 00:41:24:12

Speaker 3

And then looking at things like complications around bay roofs and the need to insulate the bay roof, when you do the external wall of ice, you really to increase the risk in that localized area. So the reports will come out and we'll discuss all of these different junctions. Essentially, you know, sometimes some of the junctions were found to be really critical.

00:41:24:12 - 00:41:41:14

Speaker 3

Yes. You have to pay a lot of attention to this specific area of the times. It's a little bit more flexibility. And if you're managing the risk in practice, you wouldn't need to discern certain things. And I think more guidance on these areas is going to be really important output and we would encourage more work in that area.

00:41:41:14 - 00:42:06:06

Speaker 3

Okay. But that's only one way of assessing one risk and there's lots more things that we would need to do. So now we'll talk about interstitial condensation risk from IWI, which probably is one of the most common questions that we get to look at. So this is a graphic which shows a profile of performance depending on performance in terms of energy consumption, energy savings and moisture risk.

00:42:07:03 - 00:42:24:21

Speaker 3

Energy savings is the orange line and moisture risk is the blue line according to u-value of the wall, i.e. how much insulation you've put on some on the left hand side at 1.6 U-value you haven't got very much insulation at all, if anything, and right at the other end, at point two, you've got a big chunk of insulation on that wall.

00:42:25:20 - 00:42:44:22

Speaker 3

What we wanted to do is see where the sweet spot was in terms of risk kicking in and energy savings being maintained. So at the very end, that downside very low risk, but not very much energy saving because you've not got very much insulation on. And as you move up towards where that red dotted line is at point eight, there's an inflection point here.

00:42:45:02 - 00:43:24:20

Speaker 3

So at that point you've maintained relatively robust energy savings time over the thickness, but you instantly get a kick up in the risk of interstitial conversation about point eight. So whilst you're making some energy savings, if you go beyond .8 you're making dramatic increase in risk. So I think most of us know that the building regulations is pushing people to .3 so I think the evidence that we can show here is that if you continue to push people with the IWI to .3 yes, you are marginally improving the energy performance of the wall by about 20 30%, but you are massively increasing the risk of interstitial condensation.

00:43:24:20 - 00:43:52:14

Speaker 3

And so a balance between risk and energy saving is perhaps needed and would suggest the current approach isn't necessarily balanced. This is a very identical graphic, except that this time we've stopped the external wall being able to reject its moisture to the inside and the outside. The previous graphic it could do that. This graphic can only put the moisture, it can only lose moisture to the outside.

00:43:52:17 - 00:44:07:20

Speaker 3

And what that's doing in the model is essentially saying that the risk of condensation, interstitial condensation increases because it can't get rid of the moisture as quickly. Okay. The might be some questions for that which I'm I happy to answer, I appreciate so very quick.

00:44:08:01 - 00:44:19:16

Speaker 1

I'm just going to jump in and say if you if you could wrap up in a couple of minutes, just got a bit of time to ask your questions at the end and have a chat. So we've got it's 48. So if you have two more minutes, three more minutes though I'm hoping.

00:44:19:24 - 00:44:36:07

Speaker 3

Yeah, that's fine. Yeah. Well that graph isn't even showing for some reason overheating. We can we can take overheating risks if people want to. And so the risk take homes. I knew I wouldn't get to the interesting findings. The next slide is the interesting findings. But let's stop here and you'll have to wait for the papers to come out.

00:44:36:07 - 00:45:00:07

Speaker 3

So the interesting findings. So what's the risk take homes we've got we've identified that lots of risks exist. It's very variable and it needs a lot of consideration. But and so it needs better to a survey tools and more consideration in terms of modelling because if you can actually spot it, you can do something about it. So I think, yes, there are high risk retrofits, but you can address the problems and you can create designs that mitigate the risk.

00:45:00:07 - 00:45:27:16

Speaker 3

So that was that was a really nice finding that, yes, there are risks, there increase, but you can do something about it. And I think a really important risk profile that we were looking at is that actually there's so much uncertainty in models and the application of thresholds is really not a robust way to assess risk and it's only looking at one risk and almost gives people permission to think that they don't have a risk by being over a threshold and that that's just as dangerous as being marginally below a threshold.

00:45:28:08 - 00:45:49:14

Speaker 3

One of the ways which it didn't show on the graphic was some reason it didn't appear and currently lots of assessment methods are only looking at a single year's worth of models. And really what we need to be able to say is not that this passes the threshold or fails the threshold, but we need to say something along the lines of six out of ten years.

00:45:49:14 - 00:46:14:24

Speaker 3

This fails versus two out of ten this fails. And that just gives us a little bit more understanding of the fact that risk is a continuous kind of it's a risk is a profile that doesn't have that much certainty involved and that some of the modelling that we've kind of hopefully advanced that can be picked up and embedded into future and future standards and approaches to risk assessments.

00:46:14:24 - 00:46:21:12

Speaker 3

And I can skip all the way through. And just to say, are there any any questions? No. And stop it now.

Was airtightness and draughtproofing a consideration in any of the measures undertaken? (You could feasibly insulate a suspended timber floor and actually make things worse by putting lots of holes in it)

Yes, we measured air tightness before and after every retrofit, there was no predictable trend, some retrofits achieved incidental air tightness improvements, others did not and some made it worse.

Was there any difference in the SWI results between the brick and stone buildings?

No the impact of the retrofit was determined by the insulation performance, not the wall type.

Insulating external walls externally v. insulating them internally? Assuming externally is vastly superior.

Both IWI and EWI have surface condensation risks associated with discontinuities in the insulation layer. However, IWI additionally has interstitial condensation risks. In terms of thermal performance there is no difference if they should have the same specification (U-value).

Just wondering what assessment of building condition / how damp a building was pre-retrofit and what repair works were done as to ensure that the retrofit was being installed to a dry building?

We would repair any leaks and dry the building out prior to retrofit. This often took days or weeks and cost thousands of pounds.

Different U-value measurements within the same element - did you identify the reasons? e.g. differences in areas with varying damp content?

It was not always possible to identify this, however, air movement through the building fabric, differences in construction make up and proximity to thermal bridges were all influential. Moisture can impact U-values but none of the baric elements had particularly notable moisture contents.

What was the spec / U-value target for the solid wall insulation fitted generally? Just to building regs or below / beyond?

EWI was specified to building regs, IWI was specified to manufacturers’ designs based on minimising interstitial condensation risk.

Were you assessing non-energy benefits, e.g. thermal comfort, indoor air quality, etc.?

No this would have required longitudinal monitoring in the occupied home. We were only able to test vacant homes.

SWI - joins at corners etc. and skilled/unskilled installers - there is a risk here to success, fabric, health. How can we advise people on what they should look for in specifying and commissioning works?

Good question, retrofit coordinators should ensure the designs do not have built-in discontinuities, onsite mid-installation visits with thermography or photographic evidence would be needed to ensure that products are installed correctly, plaster finishes hide discontinuities.

With the SWI did you differentiate the benefit between those houses which were terraced and those which were semi-detached and had much more walls that were heat loss elements?

Yes, this will be identified in the reports, clearly the greater the wall area insulated the greater the savings.

With the Heat Flux (U-value) measurement methodology, can this be done reliably when there is an air gap (or cavity) in the wall makeup  i.e. Internal wall insulation with gap between the IWI and brick wall), or narrow cavity? 

In theory, the measurement will provide the actual equivalent U-value, i.e., with the impact of the air movement factored in. For instance if the wall should have a U-value of 0.3 but there is air movement behind the insulation we might measure a U-value of 0.9. We measured this several times within DEEP. However, measuring a U-value also requires some degree of stability. If there is considerable air movement (due to windy weather for example) this can introduce too much variation to achieve a robust result. Long measurement periods help to mitigate against this.

What type of SWI did you use in the study - internal or external?

Both. There will be a case study report published for each home providing all the details.

Many thanks David. Are you able to comment on whether the big benefits from SWI were due to the external insulation directly, or due to the fact that the external insulation may have allowed the existing masonry to dry and consequent reduction in thermal conductivity as a result of lower moisture levels in the masonry? Another way of looking at this is whether any walls were lime rendered and washed for a comparison - this constitutes a possible control group. Also  both SWI and lime render will increase air tightness. Please comment on these aspects.

The benefit is not due to moisture content, Moisture measurements were taken and no notable differences pre- or post-retrofit were observed. Changes in air tightness after retrofit were also not usually notably different.

Did you find a difference between external wall insulation and internal wall insulation on the solid walls? Did you employ the same type of insulation in all the case studies, sorry if I missed this.

Both IWI and EWI have surface condensation risks associated with discontinuities in the insulation layer. However, IWI additionally has interstitial condensation risks. In terms of thermal performance there is no difference If they have the same specification (U-value).

Landlords - currently there is a spending cap on EPC does this encourage the piecemeal approach to retrofit, do they return to bridge the gaps between roof and wall etc?

The whole house approach is more expensive, so landlords operating outside of PAS2035 will be incentivised to follow the piecemeal approach. We hope our evidence on the increased risks of this approach will discourage this approach. Not insulating a home and helping the tenant in other ways may be better than insulating it poorly.

Have you used PHPP as a modelling tool? how does it compare with those analysed?

No but we would like to do this in a future round of research.

I think your statement that homes that don’t have wall insulation are unhealthy is completely untrue, living in a traditionally built house that is well repaired with the correct materials has minimum surface condensation risk. Living in a poorly repaired traditional house does have a high risk of surface condensation, the issue is in the plaster and paints used, not that it is uninsulated.

Agreed! I answered this on the day but to reiterate the analysis is simply saying that the majority of junctions are at risk of surface condensation (according to the temperature factor calculations), already in uninsulated homes. This doesn’t mean there will be any condensation if the home is appropriately heated and ventilated. Also the models used are subject to uncertainty so even a predicted risk may not manifest in practice.

Have I missed cavity wall insulation in the presentation? Were any of these included in the case studies? How does this compare with SWI?

Sorry the project is only focussed on solid walled homes.

When you say “high risk of surface condensation in un-insulated solid walls”, can you confirm what type of internal finishes were assessed – were they hygroscopic? Were the solid walls dry or in need of external repairs?

The homes all had gypsum plaster finishes however the risk was associated with the surface temperatures which the surface finish had a minor impact on, the surface temperature was driven by the whole wall U-value.

Over what time range were measurements taken? Benefits may be fine in the short term but could be problematic in the long term.

Measurements were taken over a period generally of 4 weeks under steady state conditions i.e., a continuous temperature of 23 degrees was generally achieved 24 hours per day. The methods for all the data analysis are described in the final reports.

On a previous webinar I asked who we should direct applicants to for this sort of research to inform their retrofit proposals and whole house approach. Professional body/specialist qualifications. Building Surveyors via the RICS/Architects through the RIBA, or environmental building surveyors?

Not sure I fully understand the context but this research should be published on the government website in November and we are happy to receive enquiries on our research at Leeds Beckett.

Are you saying the EPC is unreliable given the emphasis placed on it?

Yes, it is not designed to predict the specific impact of specific retrofits for specific homes, though this is what it has begun to be used for. Our research quantifies the extent to which it is not able to provide accurate information on a house-by-house basis, but again it is not designed to do so.

Is it recognised that using wood fibre insulation on old stone walls that have a high level of inherent damp could instigate dry rot.

In our models, the build-up of moisture in a wall post IWI insulation is determined by the thickness of insulation., regardless of product choice. Any vapour open insulation was better able to avoid moisture accumulation compared to non-vapour open alternatives.

Has any research been undertaken on the risks / benefits of cavity wall insulation where cavities are quite narrow, as commonly found on mid-century houses?

Sorry the project is only focussed on solid walled homes.

what thickness of EWI insulation can you add before it becomes too risky?  9" solid wall for example?

There are no inherent risks associated with the thickness of EWI applied since the fabric is on the warm side of EWI.  IWI however posed lower risks when installed in thicknesses that may achieve U-values up to 0.8. Any U-values below (i.e., to achieve building regulations) and this and risks increased significantly.

What is the cost of doing an individual dwelling survey prior to choosing what retrofit measures to take?

Apologies we did all the surveys ourselves using our researchers so cannot provide a response. However, we often found that without destructive investigations, air tests and thermography, surveys would not be effective at identifying performance issues or potential risks needed to inform safe designs.

Is there a plan to investigate similar interventions that are appropriate for protected historic buildings e.g. nationally listed or within conservation areas etc? EWI would be totally inappropriate for the vast majority of these....

Findings from this project can be applied to many situations in historic homes. The full reports will provide more information.

For SWI was there a difference noted between internal and external insulation?

Both IWI and EWI have surface condensation risks associated with discontinuities in the insulation layer. However, IWI additionally has interstitial condensation risks. In terms of thermal performance there is no difference If they have the same specification (U-value).

Was indoor air quality tested pre and post retrofit? Without MVHR is there not a huge risk of AQ deterioration?

No this would have required longitudinal monitoring in the occupied home. We were only able to test vacant homes.

I would be very interested in seeing an investigation of whole house retrofit using vapour permeable traditional materials e.g. insulated non-hydraulic lime renders and plasters, wool or wood based loft insulation, also secondary glazing rather than replacing original windows etc..  It would also be interesting to measure human perceptions of 'comfort'..... Is this something that is being investigated?

The project did investigate the use of vapour open materials and secondary glazing. Comfort could not be investigated as this would have required longitudinal monitoring in the occupied home. We were only able to test vacant homes.

Did your research consider if the initial retrofit surveyor was suitable qualified?

We undertook all retrofit surveys and did not employ a specialist for this role, though our team includes a qualified retrofit coordinator.