On the 20th January 2017 Simon Coveney signed S.I. 4/2017. This law is now the endgame for Ireland’s drive towards low-energy buildings, as it effectively kicks off the drive towards Nearly-Zero Energy Buildings. The Minister also began a public consultation on the 24th March, with the purpose of reviewing Ireland’s energy-efficiency regulations.
This programme is ambitious, well-intentioned and – in many ways – laudable. It is also, almost certainly, disastrous.
The temptation is to say that the move to Nearly Zero-Energy Buildings (NZEB – this piece will try to minimise the acronyms, but some are unavoidable) has planted a time-bomb under Irish houses, and the only question is when that bomb will explode. However, that isn’t quite accurate; the time-bomb was planted many years ago. NZEB legislation will just add more explosives to the cache.
This piece is a story. It’s a story about the increasing marginalisation of people and professionals from regulation; it’s the story of how power avoids scrutiny by controlling exactly who can scrutinise it; it’s about how regulation doesn’t just regulate what gets done, but also that only the right sort of people can do it. It’s also about the move to Nearly-Zero Energy Buildings, and how that move has alienated hundreds of thousands of people from a part of the construction process; it’s about how this decision is presented as a necessity, when it is not. Finally, and most importantly, it’s about why that alienation will ultimately prove disastrous.
Insulation: It’s complicated
Perhaps the most important thing to stress about insulation is that in the space of a generation, it has made houses more complex – and and more prone to failure – than they ever have been before.
Imagine a solid brick wall, like you might get in an old-fashioned Georgian house. Or, if you like, a two-foot thick rubble wall that you’d see in a farmhouse.
Those walls are put together with great skill, but the way they work is pretty simple. They get wet, either by rain from the outside, or condensation from the inside. Whenever the house is heated, the walls are gently heated too; any moisture is evaporated away and the wall dries out. Some of the evaporation will go into the house, but the house is draughty, so it will be ventilated away before there’s a problem.
Put some insulation on that wall, and you completely change how it works. External insulation stops the wall being well-ventilated; internal insulation is even more problematic, as the insulation stops the wall from ever being heated and so it may never dry out. Any water that finds its way into the wall – left over from construction, or from driving rain, or internal water vapour condensing on the cold wall, or a host of other sources – has no route out.
What you’ll end up with is the inside face of the wall being cold and damp – bluntly, you might as well build a mould incubator. So your wall will be slowly eaten away while lots of spores and fungus grow within them, and you won’t know anything’s wrong until you faintly smell damp, or you see the odd patch of mildew, or your nine-month old child has breathing difficulties that he or she can’t seem to shake and you can’t move house and there’s nothing you can do.
(Just to be clear, I’m not exaggerating. I suffer from mild asthma, and spent two years living in a generally-pretty-nice apartment in the UK. After a few months I begin to wonder why my inhaler use had increased since I moved in there; I checked the safety file, discovered it was internally insulated, and stopped wondering. Since I moved out my inhaler use has promptly fallen again, I’d guess to less than one-tenth of what it had been.)
Insulation is complicated. And the problem isn’t just that the industry has not kept pace; the problem is that the various regulatory texts haven’t kept up either. The Homebond Manual, once a decent bible for construction, still has 1990s-esque cavity wall details that should have been junked a decade ago. The SEAI grant scheme will only pay out for internal insulation details that are technically outmoded and a condensation risk. The government’s own “Acceptable Construction Details” still include internal insulation on hollow block walls – a terrible form of construction that should quite simply be banned – but they don’t include details for a door threshold or a rooflight.
In other words, “the industry hasn’t caught up” is minimising the problem. The people who are supposed to be advising and regulating the industry haven’t caught up and, if anything, some sections of the industry are ahead of them.
This problem didn’t occur when Minister Coveney signed the act into effect. It occurred over a decade ago. Start points are always a little too neat, but 2005 is as good a place as any.
Part L: It’s A Process
In 2005, a quiet revolution occurred in Ireland’s energy-efficiency regulations (The Building Regulations Part L and its associated Technical Guidance Document – I’m simply going to refer to Part L throughout this piece). Up until then, Part L had effectively been a series of “speed limits.” Building elements were all given maximum U-Values – a U-Value is a measurement of how much heat passes through a surface – and that was it. The walls of a house had to be 0.27W/m²K, the floors had to be 0.25W/m²K, and so on. These are absolute or easily-calculated values that give fixed maximums or minimums – we might call it “the value limit approach.”
The advantage of value limits is that they’re clear, unambiguous and easily understood. The disadvantage is that they’re rather crude, and don’t police everything. A house might hit all the U-Value targets, but if it’s heated entirely by electricity then it will still cost its owners a fortune. So in a new step to the procedure was introduced; as well as complying with all the minimum insulation values in Part L, the key data about the house had to be inputted into a computer program (a glorified excel spreadsheet, really). This created a virtual energy model of the house and checked how much carbon it produced, based on standardised occupancy patterns. It also checked how much carbon the house would produce if it was built using the minimum Part L values, but with the other variables (the type of energy used, the orientation of the windows, the efficiency of the boiler and so on) selected to reflect a “typical” house. The designed house was checked against this imaginary house, or Reference Dwelling. If the house produced more CO2 emissions than the Reference Dwelling, it would fail.
This had a straightforward aim – keeping new houses honest, If you like – but it was a fairly complex and roundabout way of doing it. If you took physics for the Leaving Cert, and are handy with MS Excel, you’d probably find the procedure pretty straightforward. For people who don’t fall into that bracket – and architecture, not to mention construction generally, attracts a wide range of people – it was difficult and alienating. And yet it went… well, unnoticed is too strong a word, but unremarked is about right. Ireland’s self-assessment system meant there was nothing to force anyone to actually do the calculation, and you would automatically pass in the overwhelming majority of cases anyway – it would really only be an issue if there was something very odd about the design (say, all the house’s windows were facing North, or it was entirely heated by electricity).
So in many cases, it simply wasn’t done.
This process would later become known as the Domestic Energy Assessment procedure, or DEAP – and the computer program is now the single most important specialist piece of software available to the construction industry in Ireland. This is because after 2005, three things happened which catapulted DEAP to the forefront of how the energy use of houses was calculated.
It began with the European Union initiating a drive towards low-energy houses; the policy is currently laid down in Directive 2010/31/EU. Meanwhile, the Green Party formed a government with Fianna Fáil and took these targets seriously, instituting two reforms: the first was the introduction of Building Energy Ratings, the second was a 60% improvement in Part L.
Building Energy Ratings are fairly simple in theory; they’re a measurement of how much energy a house uses per square metre of floor area. There’s a fundamental issue with the basic premise, though: a house doesn’t use any energy – it’s the people living in it who do that, and people are a fairly varied bunch. You can have a single elderly person, or a family of five, or a working couple all living in identical houses, and they will use differing amounts of energy. They will heat the house at different times, to different temperatures. Some people will heat to 22°C, some to 20°C, some will wrap up in a blanket or stick a hot water bottle under their jumper and be happy with 16°C. An approximation was needed, so the Department of the Environment decided that a typical household would heat the house eight hours daily for nine months a year, and that they’d heat the living room and kitchen to 21°C while keeping the rest of the house to 18°C. They also decided that a three-bedroom semi-detached house constituted a typical dwelling. Assessed using DEAP, this typical house seemed to use about 150 units of energy per square metre every year. So this figure became the “base” figure for BER – it corresponded to a C1 – and all the other figures ratings are based on it.
You may have noticed that the word “decided” is appearing rather more than it should. I’m using it with precision: there doesn’t seem to have been any detailed research into how much energy people actually use, or what really constitutes a typical household. These figures don’t correspond to anything in particular, and hardly anybody in the country will really occupy a house as DEAP’s standardised occupancy claims they will. People in poorly-insulated properties won’t even try to heat them to 21°C – the costs involved would be astronomical. People in low-energy homes may have a tendency to heat them more than this, because it’s relatively cheap to do so.
This doesn’t have to be an issue if the figures are used properly. BERs are only ever supposed to be thumbnail views on how energy-efficient a house is, they aren’t meant to be comprehensive predictions of how much energy you will use per annum. A BER is only meant to let a buyer or renter see that a house is F-rated and think “uh-oh, my bills will be high.”
The problem is that, having decided on these not-at-all researched, lick-a-finger-and-stick-it-in-the-air figures, the state also appear to have decided they are real.
What is a 60% Improvement, Anyway?
A 60% improvement in the energy regulations was a laudable aim. What does it mean, though? There are dozens of ways you can “improve the regulations by 60%”, and somebody in the Department decided to use what they had. Part L already measures a house’s energy use and compares it to a “typical” house, this person thought. So what if, instead of just matching that house, it has to use 60% less energy? That’s a 60% improvement, right there.
This idea was first put into practice in 2007 (with a 40% improvement), and then stepped up to 60% in 2011. It revolves around a number called an Energy Performance Coefficient, or EPC. You input the geometry and various key pieces of information about your house into DEAP. DEAP calculates two versions of your house, just as it did back in 2005; the one you want to build, and a fictional Reference Dwelling based on the values for a “typical” house built in 2005. The energy used by the Proposed Dwelling is divided by the energy used by the Reference Dwelling. The resultant number is the EPC, and the maximum permissible EPC is 0.4; in other words, your house can only use 40% as much energy as it would have done in 2005.
That explanation should immediately highlight the two big issues. First of all, it’s arbitrary – actually, let’s not mince words, Outperforming An Imaginary House Built In 2005 is extraordinarily arbitrary. The other point is often forgotten; in the space of two years, the Building Regulations went from being easily understood by any reasonably intelligent person, to a highly complex process that only a tiny percentage of people can calculate. Unlike value limits, the chances of a self-builder being able to calculate all this is more or less zero. The overwhelming majority of the professionals who design and construct houses don’t fully understand it either, and most architects in the industry tend to subcontract the calculation out to specialists. The law has become an esoteric speciality, only understood by a privileged few.
Let’s be clear; this is a serious problem. Regulations should be made to be understood wherever possible. People who don’t fully understand laws are less likely to follow them. Worse, if you deliberately make rules that the average person can’t understand, they can no longer hold the industry to account; and if you have a situation where the average individual can’t intelligently hold an industry to account, you simply make it easier for that industry to break the rules. Regulations so complex that they exclude almost everybody from understanding them will almost certainly be broken, whether inadvertently or not. They also verge on antidemocratic.
What I find even more worrying, and disappointing, is that we don’t hear many voices within construction who see this complexity as a particular problem. It’s well-known that most architects can no longer calculate whether a house complies with Part L or not, but this is dismissed with platitudes about “upskilling.” Essentially, we’re asked to believe this complexity is necessary: the pre-2005 system was far too crude, this is a much more accurate assessment, and if some people find it too complicated that’s tough – there’s no alternative. And yet I’ve never seen any substantiation of this viewpoint, and Part B – that’s the fire regulations – have managed to take the complexity of fire engineering and turn it into a series of laborious, but relatively simple rules.
So let’s test it.
Convoluted Does Not Mean Accurate
I’ve used DEAP to test the energy use of various dwelling types; they’re all 3-bedroom dwellings with the same floor area (and an assumed occupancy of four people), and identical construction standards have been applied to each one. They’re Value Limits, picked to roughly achieve compliance with Part L and simple to understand for anyone with a construction background: -
- Wall U-Value: 0.16W/m²K
- Roof & Floor U-Value: 0.12W/m²K
- Boiler Efficiency: 91% (Gas assumed)
- No Draught Lobby or Sheltered Sides
- Window U-Value and area: 25% of floor area @ 1.2W/m²K, north-facing glazing may not exceed south-facing glazing¹
- Airtightness@50Pa: 3.5ach
- Thermal Bridging y-value (assumed, based on a typical calculated figure for dwellings): 0.05W/m²K
- Ventilation: Natural ventilation. 1no extract fan per 60m² or part thereof, minimum 2no.
- 1 chimney for dwellings 80sq.m-200m², 1no for each subsequent 100m² or part thereof.
- Renewable provision: 10kWh/m²/yr thermal
- HW Cylinder volume: 2 x floor area to maximum 500l, indirect cylinder with 100mm factory insulation.
The figures that come out are as follows:-
We can see two things here. First of all, standardised values do not cause wild fluctuations in the figures. For four of the five house types, the EPC varies by just ±0.5% from the 0.4 target (the exception is the apartment, which is still only 2.75% out, and we’ll come back to that in a moment). This simply isn’t a significant enough variation to dismiss standardised values as inaccurate.
However, there is a second issue. If we compare the EPC with the actual energy use of the buildings, it’s obvious that compliance with the law is… well at this point arbitrary is too kind a word, and perverse is nearer the mark. The three buildings with the highest energy use pass the regulations, while the two buildings with the lowest energy use fail. The apartment (which fails) has the lowest energy use but the highest EPC, while the bungalow – which uses 40% more energy per year – passes. The value limits mean that the energy use varies by ±16.6% but, if we were to demand an EPC of 0.4 from each dwelling, that figure would increase further.
It’s not too strong to say this is ridiculous. It isn’t just that the complexity of the current regulations is not particularly necessary; the compliance model is actually more arbitrary than a range of value limits would be. The structure penalises apartments in particular, and that’s particularly noteworthy given the wealth of new stories about apartments being currently too expensive for developers to build².
Oh yes, and in case you’re wondering – no, the move to Nearly Zero Energy Buildings won’t improve this situation. Predictably, it will make it worse.
NZEB: No, It’s Not A Rap Collective
“Nearly zero-energy” is an almost self-parodically vague term, but thankfully the EU have defined exactly what it means. A building is “nearly zero-energy” if its energy demand is 45kWh per square metre each year. There’s also an aspiration that every house will be partly powered by renewable energy sources generated locally (this is why Part L demands that every house gets some of its energy from renewables).
Make no mistake; 45kWh/m²/yr is very low. It also creates an an obvious issue because Part L deals with Energy Performance Coefficients, not energy used per square metre.
However you might remember one of the Department’s decisions; a typical house uses 150kWh/m²/yr. 45 is 30% of 150. So, yes you’ve guessed it, the current intention is to lower the EPC from 0.4 to 0.3.
This is insanity. We have current regulations that are not understood, completely non-functional, not fit for purpose, and based on energy use that the state essentially made up. Rather than fixing this problem, or even acknowledging it exists, the state is just ramping them up further. This isn’t “improving standards,” any more than a resource teacher is “helping” an 8 year-old with their times tables by setting tests in advanced trigonometry. Lowering the EPC further will just make its perversity worse; the 108sq.m mid-floor apartment referenced in the previous section will need to achieve 35.51kWh/m²/yr to pass, while the detached bungalow will pass with 54.35kWh/m²/yr – that’s 45% higher. The notion that the buildings will be performing at 45kWh/m²/yr is simply fictional, and apartments will be penalised even more than before.
Advocates of Passivhaus may, at this point, be screaming their conclusion at the screen: dump the reference dwelling! Let’s bin all the palaver with EPCs and just set a maximum energy use of 45kWh/m²/yr. This is exactly what the EU are targeting anyway, and systems like Passivhaus also use an energy-per-square metre assessment – as a result, many advocates of Passivhaus argue that it should simply replace the regulations altogether.
This does nothing to address the question of making the law understandable and transparent; Passivhaus is slightly more complex than DEAP. Even setting that aside, there’s a very basic problem – energy per square metre is measuring the wrong thing. To repeat what was said at the start: houses don’t use energy, people use energy. It doesn’t matter how much energy each square metre of house is using, it’s the energy use per person that counts.
The house types above make the problem clear (the same value limits have been applied as before, this time to detached houses of different sizes). For the same reason that a mug of tea will cool a lot slower than a thimbleful, larger houses use less energy per square metre than smaller ones – that’s basic physics. What this ignores is that large houses tend not to be fully occupied – so the largest of the houses shown has the lowest energy use per square metre, but it also has the highest energy use per person. Again, insisting all these houses meet a lower target will only accentuate that issue.
What needs to be understood is that Passivhaus is a comfort standard, not an energy standard. The Building Regulations are supposed to be about climate change and energy security, not about ensuring that wealthy people can keep their houses and 20°C all day, all year round. We’re back where we were with the Reference Dwelling and EPC: a hugely complicated assessment model that actually doesn’t deliver an accurate solution. Or if you prefer, a mess.
Accuracy Is 73.3% Overrated
This piece opened by comparing the pre-2005 Part L to a series of speed limits. It’s worth considering how speed limits actually work.
We’re so accustomed to them that it’s easy forget; speed limits are a very inexact science. They take no account of climactic conditions, or driver ability, or how busy a road might be at any one time. Let’s imagine for a moment that to achieve a more accurate approach to road safety, we replaced speed limits them with a procedure. Let’s say that “base” speed limits are applied to M, N, R and L-roads, and drivers were assessed periodically to give them a Driver Ability Coefficient based on their reaction times, and they were then expected to multiply “base” speed limits by their own ability coefficient, plus a further series of coefficients based on local conditions (rain, built-up areas and so on), to arrive at a personalised speed limit that would be more accurate…
Yes, OK, I’ll stop. That approach would be self-evidently absurd. Most drivers wouldn’t be able to follow the requirements: drivers just want to get from A to B, not become experts in risk assessment. It also requires so much knowledge that it becomes entirely unpoliceable.
In short, it’s a ridiculous approach… and yet that is exactly what Part L has done. Somewhere along the line, somebody forgot that most people involved in the construction industry do not particularly care about low-energy building. Whether you see this as a problem is irrelevant – we can’t trade our construction sector for Switzerland’s, so we have to work with what we’ve got.
The developers want a profit; the builders want to get it done and move on to the next thing; the architects want it to work well and look nice; the structural engineers want it to stand up. If you make it easy for these people to comply with the law, they will do it; if you give them simple rules to follow, and the resources to help them follow the rules, they’ll follow them. On the other hand, If you make laws needlessly complicated, if you drive the industry towards the use of specialists, if you utter mealy-mouthed aspirations about “upskilling the industry” while you make regulations ever-more convoluted – well, then you simply end up with a world of non-compliance. You’ll have an industry that doesn’t know how to follow the regulations, and a general public who can’t possibly hold them to account. The regulatory bodies are behaving like a Dickensian factory-owner, forcing the smallest children to carry heavier and heavier loads, outwardly ignoring their protests, already preparing to point the finger of blame at those children when they inevitably collapse.
2020 isn’t far away and there’s only two years left to sort this mess out. Vague platitudes about reforming the industry will not solve the problem. It’s time to be pragmatic.
A return of value limits will – at a stroke – make regulations easy to follow. They’re no less “accurate” than the current situation and will empower everybody to build within the law, without recourse to specialists.
DEAP should be retained as an alternative, so that there’s the same flexibility that exists today, but the EPC needs to be dumped.
The Homebond Manual needs to be re-released, hugely expanded to take into account full-fill cavity wall, timber frame and external wall insulation, with details that reflect today’s need for airtightness and insulation levels.
The Acceptable Construction Details need to be amended, so that they include common details that aren’t currently included, and don’t include constructions that should no longer be in use.
A voluntary low-energy construction accreditation needs to be introduced by the state, so that the industry is incentivised to understand this stuff, and so that builders have somewhere to go where they can actually get clear, up-to-date knowledge.
That’s a lot to do in two years. But the alternative is to continue down the road we’re on now; the result will be bad environments, bad buildings, and millions wasted when defective insulation is removed from defective walls to reveal blockwork black with mould.
A lot of this has already gone wrong, and there’s only two years to put it right. For the love of god, let’s start now. Take part in that public consultation. Send in a submission. It’s long past time we shouted stop.Endnotes 1: All windows in the buildings shown have been given an east-west orientation. 2: (Interestingly, when the Department of the Environment went to war with Dublin City Council because their guidelines had made apartments too expensive – this after developers had been furiously lobbying the government to reduce standards so they could bring down their construction costs – this issue did not come up, even though it adds thousands to apartment prices. You can draw whatever conclusions you like from this, but suffice it to say that those of us who believe “almost nobody in construction actually understands the regulations properly” were not at all surprised.